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.
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 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.
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
Contents
VI
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.
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 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
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
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.
5-16. 5-17. 5-18.
147
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
VII
viii
Contents
5-19. Effect of R-Ratio and Test Temperature on Crack Propagation of HSLA Steel Grade I 165 5-20. Effect of Test Temperature on Fatigue Crack Propagation Behavior for Two HSLA Steel Grades 166 5-21. Stress-Cycle Curves for Weldments of Different HSLA Steel Grades 167 5-22. Weldments (FCA W): SAE 980 X Steel vs 1006 168 5-23. Weldments (TIG): DOMEX 640 XP Steel Welded Joints vs Parent Metal 169 5-24. Weldments (FCAW Dressed by TIG): Fatigue Life Estimates Compared With 170 Experimental Data for SAE 980 X Steel 5-25. SAE 980 X Steel Weldment (FCAW): Smooth Specimen vs TIG-Dressed vs As-Welded 171 5-26. SAE 980 X Steel Weldment (FCAW): Lap-Shear Joints 172 5-27. Microalloyed HSLA Steels: Properties of Fusion Welds 173 5-28. Microalloyed HSLA Steels: Properties of Spot Welds 174
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
8-9. Type 304 Stainless Steel: Fatigue Crack Growth Rate at Room and Subzero Temperatures 205 8-10. Types 304 and 304L Stainless Steel: Effect of Cryogenic Temperatures on Fatigue Crack Growth Rate 206 8-11. Type 304 Stainless Steel: Fatigue Crack Growth Rate in Air With Variation in Waveforms 207 8-12. Type 304 Stainless Steel: Effect of Hold Time on Cycles to Failure 208 8-13. Type 304 Stainless Steel: Effect of Hold Time and Continuous Cycling on Fatigue Crack Growth Rates 209 8-14. Type 304 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate 210 8-15. Type 304 Stainless Steel: Effect of Frequency on Fatigue Crack Growth Behavior 211 8-16. Type 304 Stainless Steel Welded With Type 308: Fatigue Crack Growth Rates 212 8-17. Types 304 and 310 Stainless Steel: Effect of Direction on S-N 213 8-18. Types 304, 316, 321, and 348 Stainless Steel: Effects of Temperature on Fatigue Crack Growth Rates 214 8-19. Type 309S Stainless Steel: Effect of Grain Size on Fatigue Crack Growth Rate 215 8-20. Type 310S Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate 216 8-21. Type 316 Stainless Steel: Growth Rate of Fatigue Cracks in Weldments 217 8-22. Type 316 Stainless Steel: Fatigue Crack Growth Rates-Aged vs Unaged 218 8-23. Type 316 Stainless Steel: Fatigue Crack Growth Rates-Effect of Aging 219 8-24. Type 316 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate 220 8-25. Type 316 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate 221 8-26. Type 316 Stainless Steel: Fatigue Crack Growth Rate in the Annealed Condition 222 8-27. Type 316 Stainless Steel: Effect of Environment (Sodium, Helium, and Air) on Cycles to Failure 223 8-28. Types 316 and 321 Stainless Steel: Effects of Gaseous Environments on Fatigue Crack Growth Rates 224 8-29. Type 32I Stainless Steel: Effect of Hold Time on Fatigue Crack Growth Rates 225 8-30. Type 403 Stainless Steel: Effect of Environment on Fatigue Crack Growth Rate 226 8-3I. Type 403 Modified Stainless Steel: Scatter of Fatigue Crack Growth Rates 227 8-32. Type 422 Stainless Steel: Fatigue Crack Growth Rates in Precracked Specimens 228 8-33. Type 422 Stainless Steel: Fatigue Strength-Longitudinal vs Transverse 229 8-34. Type 422 Stainless Steel: Effect of Temperature on Fatigue Strength 230 8-35. Type 422 Stainless Steel: Effects of Delta Ferrite on Fatigue Strength 231 8-36. 17-4 PH Stainless Steel: Fatigue Crack Growth Rates in Airvs Salt Solution 232 8-37. 15-5 PH Stainless Steel: Fatigue Crack Growth Rates in Air vs Salt Solution 233 8-38. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Room Temperature 234 8-39. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates in Air and Sump Tank Water 235 8-40. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Subzero Temperatures 236 8-41. PH 13-8 Mo Stainless Steel: Constant-Life Fatigue Diagram 237 8-42. Types 600 and 329 Stainless Steel: S-NCurves for Two Processing Methods 238 8-43. Grade 21-6-9 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rates 239 8-44. Kromarc 58 Stainless Steel: Effect of Cryogenic Temperatures on Weldments 240 8-45. Pyromet 538 Stainless Steel: Effects of Welding Methods on Fatigue Crack Growth Rates 241 8-46. Duplex Stainless Steel KCR 171: Corrosion Fatigue 242
ix
x
Contents
SECTION 9: Maraging Steels
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
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
xi
xii
Contents
12-10. Alloy 2024-T4: High-Cycle vs Low-Cycle Fatigue 328 12-11. Alloy 2024-T4: Relationship of Stress and Fatigue Cycles 329 12-12. Alloy 2024-T4: Dependence of the Average Rocking Curve Halfwidth 7J on Distance From the Surface 330 12-13. Alloys 2024 and X2024: Effect of Alloy Purity on Cycles to Failure 331 12-14. Alloys 2024 and 2124: Relationship of Particle Size and Fatigue Characteristics 332 12-15. Alloys 2024-T4 and 2124-T4: Comparison of Resistance to Fatigue Crack Initiation 333 12-16. Alloys 2024-TJ and 7075-T6: Summary of Fatigue Crack Growth Rates 334 12-17. Alloys 2024-T4 and 7075-T6: Effect of Product Form and Notches 335 12-18. Alloys 2024-T351 and 7075-T73XXX: Comparison of P / M Extrusions and Rod 336 12-19. Alloy 2048-T851: Longitudinal vs Transverse for Axial Fatigue 337 12-20. Alloy 2048-T851: Notched vs Unnotched Specimens at Room and Elevated Temperatures 338 12-21. Alloy 2048-T851: Fatigue Crack Propagation Rates in LT and TL Orientations 339 340 12-22. Alloy 2048-T85I: Modified Goodman Diagram for Axial Fatigue 12-23. Alloy 2219-T851: Dependence of Relaxation Behavior on the Cyclic Hardening Parameter 341 12-24. Alloy 2219-T851: Effect of Strain Amplitude on the Relaxation of Residual Surface Stress With Fatigue 342 12-25. Alloy 2219-T851: Relationship of Fatigue Cycles to Different Depth Distributions of Surface Stress 343 12-26. Alloy 2219-T851: Probability of Fatigue Failure 344 12-27. Alloys 3003-0, 5154-H34 and 6061-T6: Effect of Alloy on Fatigue Characteristics of Weldments 345 12-28. Alloy 5083-0 Plate: Effect of Orientation on Fatigue Crack Growth Rates 346 12-29. Alloy 5083-0 Plate: Effect of Temperature and Humidity on Fatigue Crack Growth Rates 347 12-30. Alloys 5086-H34, 5086-H36, 6061-T6, 7075-T73 and 2024-T3: Comparative Resistance to Axial-Stress Fatigue 348 12-31. Alloys 5083-0/5183: Fatigue Life Predictions and Experimental Data Results for Double V-Butt Welds 349 12-32. Alloys 5083-0/5183: Predicted Effect of Stress Relief and Stress Ratio on Fatigue Life of Butt Welds 350 12-33. 7XXX Alloys: Cyclic Strain vs Crack Initiation Life 351 12-34. Alloy 7050: Influence of Alloy Composition and Dispersoid Effect on Mean Calculated Fatigue Life 352 353 12-35. Alloy 7050: Effect of Grain Shape on Cycles to Failure 12-36. Alloy 7075 (TMP, T6 and T651): Effect of Thermomechanical Processing on Cycles to Failure 354 12-37. Alloys 7075 and 7475: Effect of Inclusion Density on Cycles to Failure 355 356 12-38. Alloy 7075: Effect of TMT on Cycles to Failure 12-39. Alloys 7075 and 7050: Relative Ranking for Constant Amplitude and Periodic Overload 357 12-40. Alloy 7075: Effect of Environment and Mode of Loading 358 12-41. Alloy 7075-T6: Effects of Corrosion and Pre-Corrosion 359 360 12-42. Alloy 7075-T73: Effect of a 3.5% NaCl Environment on Cycles to Failure 12-43. Alloy 7075: Effect of Cathodic Polarization on Fatigue Behavior 361 12-44. Alloy 7075-T6: Effect of Surface Treatments and Notch Designs on Number of Cycles to Failure 362 364 12-45. Alloy 7075-T6: Effect of R-Ratio on Fatigue Crack Propagation 12-46. Alloy 7075: Effect of Predeformation on Fatigue Crack Propagation Rates 365 12-47. Alloys 7075 and 2024-T3: Comparative Fatigue Crack Growth Rates for Two Alloys in Varying Humidity 366 367 12-48. Alloy 7075-T65I: Fatigue Life as Related to Harmonic Generation 12-49. Alloys 7075-T6 and 7475-T73: Effect of Laser-Shock Treatment on Fatigue Properties 368 12-50. Alloy 7075-T6: Effect of Laser-Shock Treatment on Hi-Lok Joints 369
Contents
12-51. 12-52. 12-53. 12-54. 14-55. 12-56. 12-57. 12-58. 12-59. 12-60. 12-61. 12-62. 12-63. 12-64. 12-65. 12-66. 12-67. 12-68. 12-69. 12-70. 12-71. 12-72. 12-73. 12-74. 12-75. 12-76.
Alloy 7075 (High Purity): Effect of Iron and Silicon on Cycles to Failure 370 Alloy X-7075: Effect of Grain Size on Cycles to Failure 371 Alloy X-7075: Effect of Grain Size on Stress-Life Behavior 372 Alloy X-7075: Effect of Environment; Air vs Vacuum 373 Alloy X-7075: Effect of Environment on Two Different Grain Sizes 374 Alloy X-7075: Effect of Grain-Boundary Ledges on Cycles to Failure 375 Alloys X-7075 and 7075: Effects of Chromium Inclusions on Fatigue Crack Propagation 376 Alloy 7475-T6: S-N Diagram for a Superplastic Fine-Grain Alloy 377 Alloy 7475: Effect of Alignment of Grain Boundaries on Cycles to Failure 378 Alloy 7475-T6: Superplastic vs Nonsuperplastic, as Related to Fatigue Crack Growth 379 Alloys X-7075 and 7075: Effect of Chromium-Containing Inclusions on Cycles to Failure 380 Aluminum Forging Alloys: Stress Amplitude vs Reversals to Failure 381 AI-5Mg-0.5Ag: Effect of Condition on Fatigue Characteristics 382 AI-Zn-Mg and AI-Zn-Mg-Zr: Effect of Grain Size on Strain-Life Behavior 383 AI-Zn-Mg: Strain-Life Curves of a Large-Grained Alloy 384 Aluminum With a Copper Overlay: Stress Amplitude vs Cycles to Failure 385 P/M Alloys 7090 and 7091 vs Extruded 2024 386 P / M Alloys 7090 and 709I vs 1/ M 7050 and 7075 Products 387 P/M Aluminum Alloys: Typical Fatigue Behavior 388 P / M Aluminum Alloys: Comparison With Specimens Made by Ingot Metallurgy 389 P/M Aluminum Alloys: Comparison With Forged 7175 for Cycles to Failure 390 Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack Growth Rates; Air vs Salt Water 391 Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack Growth Rates in Salt Water 392 Various Aluminum Alloys: Wrought vs Cast, and Influence of Casting Method on Fatigue Life 393 Aluminum Casting Alloy AL-195: Interrelationship of Fatigue Properties With Degree of Porosity 394 Aluminum Casting Alloy LM25-T6: Squeeze Formed vs Chill Cast; Effect on Reversals to Failure 395
SECTION 13: Copper Alloys
396
13-1. Copper: Effect of Air and Water Vapor on Cycles to Failure 396 13-2. Copper: Applied Plastic-Strain Amplitude vs Fatigue Life 397 13-3. Copper Alloy CI 1000 (ETP Wire): Effect of Temperature on Fatigue Strength 398 13-4. Copper Alloy C26000 (Cartridge Brass): Influence of Grain Size and Cold Work on Cycles to Failure 399 13-5. Copper Alloy C83600 (Leaded Red Brass): S-N Curves; Scatter Band 400 13-6. Copper Alloy C86500 (Manganese Bronze): S-N Curves; Scatter Band 401 13-7. Copper Alloys C87500 and C87800 (Silicon Brasses): S-N Curves; Scatter Band 402 13-8. Copper Alloy C92200 (Navy "M" Bronze): S-N Curves; Scatter Band 403 13-9. Copper Alloy C93700 (High-Leaded Tin Bronze): S-NCurves; Scatter Band 404 13-10. Copper Alloy No. 192: Effect of Salt Spray on Tubes 405 13-1 I. Copper Alloy 955: Goodman-Type Diagram 406
SECTION 14: Magnesium Alloys
407
14-1. Magnesium Casting Alloy QE22A-T6: Effects of Notches and Testing Temperature 407 14-2. Magnesium Casting Alloy QH2 IA-T6: S- N Curves; Effects of Notches and Testing Temperature 408 14-3. Mg-AI-Zn Casting Alloys: Effects of Surface Conditions on Fatigue Properties 409
xiii
xiv
Contents
SECTION 15: Molybdenum
410
15-1. Molybdenum: Fatigue Limit Ratio vs Temperature
SECTION 16: Tin Alloys
410
411
16-1. Tin-Lead Soldering Alloy: S-N Data for Soldered Joints 411 16-2. Babbitt: Variation of Bearing Life With Babbitt Thickness 412 16-3. SAEI2 Bearing Alloy: Effect of Temperature on Fatigue Life 413
SECTION 17: Titanium and Titanium Alloys
414
17-1. Unalloyed Titanium, Grade 3: S-N Curves for Annealed vs Cold Rolled 414 17-2. Unalloyed Titanium, Grade 4: S-N Curves for Three Testing Temperatures 415 17-3. Ti-24V and Ti-32V: Stress Amplitude vs Cycles to Failure 416 17-4. Ti-5AI-2.5Sn: Effects of Notches and Types of Surface Finish 417 17-5. Ti-5AI-2.5Sn and Ti-6AI-4V: Fatigue Crack Growth Rates 418 17-6. Ti-6AI-6V-2Sn: Effects of Machining and Grinding 419 17-7. Ti-6AI-6V-2Sn (HIP): S-N Curves for Titanium Alloy Powder Consolidated by HIP 420 17-8. Ti-6AI-6V-2Sn (HIP): S-N Curves for Annealed Plate vs HIP 421 17-9. Ti-6AI-2Sn-4Zr-2Mo: Bar Chart Presentation on Effects of Machining and Grinding 422 17-10. Ti-6AI-2Sn-4Zr-2Mo: Constant-Life Fatigue Diagram 423 17-11. Ti-6AI-2Sn-4Zr-6Mo: Low-Cycle Axial Fatigue Curves 424 17-12. Ti-8Mo-2Fe-3AI: S-NCurves; 'Solution Treated and Aged Condition 425 17-13. Ti-IOV-2Fe-3AI: S-N Curves; Notched vs Unnotched Specimens in Axial Fatigue 426 17-14. Ti-IOV-2Fe-3AI and Ti-6AI-4V: Comparison of Fatigue Crack Growth Rates 427 17-15. Ti-IOV-2Fe-3AI: S-N Curve; Notched Bar Fatigue Life for a Series of Forgings Compared With Ti-6AI-4V Plate 428 17-16. Ti-I3V-IICr-3AI: Constant-Life Fatigue Diagrams 429 17-17. Ti-6AI-4V: Effect of Condition and Notches on Fatigue Characteristics 430 17-18. Ti-6AI-4V: Effect of Direction on Endurance 431 17-19. Ti-6AI-4V: Effect of Isothermally Rolled vs Extruded Material on Cycles to Failure 432 17-20. Ti-6AI-4V: Comparison of Wrought vs Isostatically Pressed Material for Cycles to Failure 433 17-21. Ti-6AI-4V: Effect of Fretting and Temperature on Cycles to Failure 434 17-22. Ti-6AI-4V (Beta Rolled): Effect of Finishing Operations on Cycles to Failure 435 17-23. Ti-6AI-4V: Effect of Yield Strength on Stress-Life Behavior 436 17-24. Ti-6AI-4V: Effect of Stress Relief on Cycles to Failure 437 17-25. Ti-6AI-4V: Interrelationship of Machining Practice and Cutting Fluids on Cycles to Failure 438 17-26. Ti-6AI-4V: Relative Effects of Machining and Grinding Operations on Endurance Limit 439 17-27. Ti-6AI-4V: Effects of Various Metal Removal Operations on Endurance Limit 440 17-28. Ti-6AI-4V: Effect of Texture on Fatigue Strength 441 17-29. Ti-6AI-4V: Effect of Complex Texture on Cycles to Failure 422 17-30. Ti-6AI-4V: Effect of Texture and Environment on Cycles to Failure 443 17-31. Ti-6AI-4V: Fatigue Crack Growth Rates 444 17-32. Ti-6AI-4V: Fatigue Crack Growth Rates for ISR Tee, and Extrusions 445 17-33. Ti-6AI-4V: Fatigue Crack Growth Rates 446 17-34. Ti-6AI-4V: Effect of Final Cooling on Fatigue Crack Growth Rates 447 17-35. Ti-6AI-4V: Effect of Dwell Time on Fatigue Crack Growth Rates 448 17-36. Ti-6AI-4V: Fatigue Crack Growth Data 449 17-37. Ti-6AI-4V P / M: Comparison of HIP'd Material With Alpha-Beta Forgings for Cycles to Failure 450
xv
Contents
17-38. Ti-6AI-4V PI M: Comparisons of HIP'd Material With Annealed Plate for Cycles to Failure 45 I 452 17-39. Ti-6AI-4V P/M: Effect of Powder Mesh Size on Fatigue Properties 17-40. Ti-6AI-4V P/M: Comparison of Blended Elemental, Prealloyed and Wrought Material for Effect on Cycles to Failure 453 454 17-41. Ti-6AI-4V: P/M Compacts vs 11M Specimens: Cycles to Failure 17-42. Ti-6AI-4V: Comparison of Specimens Processed by Various Fabrication Processes for Cycles to Failure 455 456 17-43. Ti-6AI-4V: Comparison of Fatigue Crack Growth Rate, PI M vs II M 17-44. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure 457 17-45. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure 458 17-46. Ti-6AI-4V EB Weldments: Base Metal Compared With Flawless Weldments 459 17-47. Ti-6AI-4V EB Weldments: Effects of Porosity on Cycles to Failure 460 17-48. Ti-6AI-4V Gas Metal-Arc Weldments: Effects of Porosity on Cycles to Failure 461 17-49. Ti-6AI-4V: Unwelded vs Electron Beam Welded Material for Cycles to Failure 462 463 17-50. Ti-6AI-4V: S-N Diagram for Laser-Welded Sheet 464 17-51. Ti-6AI-4V (Cast): S-N Diagram for Notched Specimens
SECTION 18: Zirconium
465
18-1. 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 3,4 and 5, on carbon and alloy steels.)
19-1. Steel Castings (General): Effect of Design and Welding Practice on Fatigue Characteristics 466 19-2. Steel Castings (General): Effects of Discontinuities on Fatigue Characteristics 467
SECTION 20: Closed-Die Forgings
468
(See also under specific grades of alloys.)
20-1. Closed-Die Steel Forgings: Effect of Surface Condition on Fatigue Limit
SECTION 21: Powder Metallurgy Parts
468
469
(See also under specific alloys.) 21-1. P/M: Relation of Density to Fatigue Limit and Fatigue Ratio
469 470 21-2. PI M: Relation of Fatigue Limit to Tensile Strength for Sintered Steels 21-3. PI M (Nickel Steels): As-Sintered vs Quenched and Tempered for Cycles to Failure 471 2 I-4. PI M (Nickel Steels): Relation Between Fatigue Limit and Tensile Strength for Sintered Steels 472 21-5. P/M (Nickel Steels): Effect of Notches on Cycles to Failure for the As-Sintered Condition 473 21-6. PI M (Nickel Steels): Effect of Notches on Cycles to Failure for the Quenched and Tempered Condition 474 21-7. P/M (Low-Carbon, 1-5%Cu): Effects of Notches and Nitriding on Cycles to Failure 475 2 I-8. PI M (Sintered Iron, Low-Carbon, No Copper): Effect of Density and Nitriding on Cycles to Failure 476 21-9. P/M: Effect of Nitriding on Ductile Iron and Sintered Iron (3%Cu) for Cycles to Failure 477
SECTION 22: Composites
478
22-1. Brass/ Mild Steel Composite: Comparison of Brass-Clad Mild Steel With Brass and Mild Steel for Cycles to Failure 478 22-2. Stainless Steell Mild Steel Composite: Comparison of Stainless-Clad Mild Steel With Stainless Steel and Mild Steel for Cycles to Failure 479
xvi
Contents
SECTION 23: Effects of Surface Treatments
480
23-1. Carbon and Alloy Steels (Seven Grades): Effects of Nitrocarburizing on Fatigue Strength 480 23-2. Carbon and Alloy Steels (Seven Grades): Effects of Tufftriding on Fatigue Characteristics 481 23-3. Carbon and Alloy Steels (Six Grades): Effects of Nitriding on Fatigue Strength 482 23-4. Carbon-Manganese Steel: Effects of Nickel Coating on Fatigue Strength 483
SECTION 24: Test Results for Component Parts 484 24-1. Coil Springs, Music Wire (Six Sizes): Data Presented by Means of a Goodman Diagram 484 24-2. Coil Springs: S-N Data for Oil-Tempered and Music Wire Grades 485 24-3. Coil Springs: Effects of Shot Peening on Cycles to Failure 486 24-4. Coil Springs, 8650 and 8660 Steels: Relation of Design Stresses and Probability of Failure 487 . 24-5. Coil Springs, HSLA Steels: Effects of Corrosion on Cycles to Failure 488 24-6. Leaf Springs, 5160 Steel: Maximum Applied Stress vs Cycles to Failure 489 24-7. Front Suspension Torsion Bar Springs, 5160H Steel: Distribution of Fatigue Results for Simulated Service Testing 490 24-8. Gears, Carburized Low-Carbon Steel: Relation of Life Factor to Required Life 491 24-9. Gears, Carburized Low-Carbon Steel: Bending Stress vs Cycles to Failure 492 24-10. Gears, Carburized Low-Carbon Steel: Effect of Shot Peening on Cycles to Failure 493 24-11. Gears, Carburized Low-Carbon Steel: Probability-Stress-Life Design Curves 494 24-12. Gears, 8620H Carburized: Bending or Contact Stress vs Cycles to Fracture or Pitting 495 24-13. Gears, 8620H Carburized: A Weibull Analysis of Bending Fatigue Data 496 24-14. Gears, 8620H Carburized: T-N Curves for Six-Pinion, Four-Square Tests 497 24-15. Hypoid Gears, 8620H Carburized: Minimum Confidence Level; Stress vs Cycles to Rupture 498 24-16. Hypoid, Zero I and Spiral Bevel Gears, 8620H Carburized: S-NScatter Band and Minimum Confidence Level 499 24-17. Spiral Bevel and Zero I Bevel Gears, 8620H Carburized: S-N Scatter Band and Minimum Confidence Level 500 24-18. Gears, 8620H Case Hardened: Relation of Life Factor to Cycles to Rupture 501 24-19. Bevel Gears, Low-Carbon Steel Case Hardened: Relation of Life Factor to Cycles to Rupture for Various Confidence Levels 502 24-20. Gears, AMS 6265: S-N Data for Cut vs Forged 503 24-21. Spur Gears, 8620H: S-N Data for Cut vs Forged 504 24-22. Gears and Pinions: PIM 4600V vs 4615; Weibull Distributions 505 24-23. Gears and Pinions: PIM Grades 4600V and 2000 vs 4615; Percent Failure vs Time 506 24-24. Gear Steel AMS 6265: Parent Metal vs Electron Beam Welded 507 24-25. Gears, 42 CrMo4 (German Specification): S-N Curves for Various Profiles 508 24-26. Gears, 42 CrMo4 (German Specification): Endurance Test Results in the Weibull Distribution Diagram 509 24-27. Bolts, 1040 and 4037 Steels: Maximum Bending Stress vs Number of Stress Cycles 510 24-28. Bolts: S-N Data for Roll Threading Before and After Heat Treatment 511 24-29. Power Shafts, AMS 6382 and AMS 6260: Electron Beam Welded vs Silver Brazed Joints 512 24-30. Axle Shafts, 1046, 1541 and 50B54 Steels: S-N Data for Induction Hardening vs Through Hardening 513 24-31. Steel Rollers, 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 514
Contents
24-32. Steel Rollers, 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 515 24-33. Linkage Arm, Cast Low-Carbon Steel: Starting Crack Size vs Cycles to Failure 516 24-34. Notched Links, Hot Rolled Low-Carbon Steel: S-N Data for Component Test Model 517 24-35. Fuselage Brace, Ti-6AI-6V-2Sn: Fatigue Endurance of HIP-Consolidated Powder 518
xvii
1
Fatigue Testing Introduction Fatigue is the progressive, localized, permanent structural change that occurs in materials subjected to fluctuating stresses and strains that may result in cracks or fracture after a sufficient number of fluctuations. Fatigue fractures are caused by the simultaneous action of cyclic stress, tensile stress and plastic strain. If anyone of these three is not present, fatigue cracking will not initiate and propagate. The cyclic stress starts the crack; the tensile stress produces crack growth (propagation). Although compressive stress will not cause fatigue, compression load may do so. The process of fatigue consists of three stages: • Initial fatigue damage leading to crack nucleation and crack initiation • Progressive cyclic growth of a crack (crack propagation) until the remaining un cracked cross section of a part becomes too weak to sustain the loads imposed • Final, sudden fracture of the remaining cross section Fatigue cracking normally results from cyclic stresses that are well below the static yield strength of the material. (In low-cycle fatigue, however, or if the material has an appreciable work-hardening rate, the stresses also may be above the static yield strength.) Fatigue cracks initiate and propagate in regions where the strain is most severe. Because most engineering materials contain defects and thus regions of stress concentration that intensify strain, most fatigue cracks initiate and grow from structural defects. Under the action of cyclic loading, a plastic zone (or region of deformation) develops at the defect tip. This zone of high deformation becomes an initiation site for a fatigue crack. The crack propagates under the applied stress through the material until complete fracture results. On the microscopic scale, the most important feature of the fatigue process is nucleation of one or more cracks under the influ-
ence of reversed stresses that exceed the flow stress, followed by development of cracks at persistent slip bands or at grain boundaries.
Prediction of Fatigue Life The fatigue life of any specimen or structure is the number of stress (strain) cycles required to cause failure. This number is a function of many variables, including stress level, stress state, cyclic wave form, fatigue environment, and the metallurgical condition of the material. Small changes in the specimen or test conditions can significantly affect fatigue behavior, making analytical prediction of fatigue life difficult. Therefore, the designer may rely on experience with similar components in service rather than on laboratory evaluation of mechanical test specimens. Laboratory tests, however, are essential in understanding fatigue behavior, and current studies with fracture mechanics test specimens are beginning to provide satisfactory design criteria. Laboratory fatigue tests can be classified as crack initiation or crack propagation. In crack initiation testing, 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. In crack propagation testing, fracture mechanics methods are used to determine the crack growth rates of preexisting cracks under cyclic loading. Fatigue crack propagation may be caused by cyclic stresses in a benign environment, or by the combined effects of cyclic stresses and an aggressive environment (corrosion fatigue).
Fatigue Crack Initiation Most laboratory fatigue testing is done either with axial loading, or in bending, thus producing only tensile and compressive stresses. The stress usually is cycled either between a maximum and a minimum tensile stress, or between a maximum tensile stress and a maximum compressive stress.
2
Fatigue Testing 1100
The latter is considered a negative tensile stress, is given an algebraic minus sign, and therefore is known as the minimum stress. The stress ratio is the algebraic ratio of two specified stress values in a stress cycle. Two commonly used stress ratios are the ratio, A, of the alternating stress amplitude to the mean stress (A = Sal S m) and the ratio, R, of the minimum stress to the maximum stress (R = Sminl
.;
Smax)'
E
If the stresses are fully reversed, the stress ratio R becomes -1; if the stresses are partially reversed, R becomes a negative number less than 1. If the stress is cycled between a maximum stress and no load, the stress ratio R becomes zero. If the stress is cycled between two tensile stresses, the stress ratio R becomes a positive number less than 1. A stress ratio R of 1indicates no variation in stress, making the test a sustained-load creep test rather than a fatigue test. Applied stresses are described by three parameters. The mean stress, S m' is the algebraic average of the maximum and minimum stresses in one cycle, S m = (S max + Smin) / 2. In the completely reversed cycle test, the mean stress is zero. The range of stress, S" is the algebraic difference between the maximum and minimum stresses in one cycle, S, = Smax - Smin' The stress amplitude, S a' is one half the range of stress, Sa = S,/ 2 = (Smax - Smin)/2.
During a fatigue test, 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, and Sa is the alternating stress, which is equal to half the stress range. Nomenclature to describe test parameters involved in cyclic stress testing are shown in Fig. 1.
Smax I---~:------;...----r---,----
Time
Ol-\-.........,~-----Ic-I-'--~r--f--+----
Smin
L..-
--""-
......._ - ' -_ _
Fig. 1 Nomenclature to describe test parameters involved in cyclic stress testing
S-N Curves. The results offatigue crack initiation tests usually are plotted as maximum stress, minimum stress, or stress amplitude to number of cycles, N, to failure using a logarithmic scale for the number of cycles. Stress is plotted on either a linear or a logarithmic scale. The result-
I
I 150 2340 steel o~ '"' 48 HRC ~ (unnotchedl'r
1000
900
:2:
800
a..
700
Q)
600
u
0.
~
U5
500 400 300 200
=
125 'iii
1::-- - - -
~
~
f-
_.
'1...
.:.!.
1 --
Fatigue limit Sf
2340 steel-= 100 Q) 48 HRC _ u (notched) .~
-cr....
Fatigue limit, Sf
l Aluminum alloy """Q; ~ ,}075-T6
j-a~-
75
0.
E
50
~,
U5 25
Stress ratio (R) = - 1
Number of cycles to fracture, N Fig. 2 Typical S-N curves for constant amplitude and sinusoidal loading
ing plot of the data is an S-N curve. Three typical S-N curves are shown in Fig. 2. The number of cycles of stress that a metal can endure before failure increases with decreasing stress. For some engineering materials such as steel (see Fig. 2) and titanium, the S-N curve becomes horizontal at a certain limiting stress. Below this limiting stress, known as the fatigue limit or endurance limit, the material can endure an infinite number of cycles without failure. Fatigue Limit and Fatigue Strength. 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, Sf' Most nonferrous metals do not exhibit a fatigue limit. Instead, their S-N curves continue to drop at a slow rate at high numbers of cycles, as shown by the curve for aluminum alloy 7075-T6 in Fig. 2. For these types of metals, fatigue strength rather than fatigue limit is reported, which is the stress to which the metal can be subjected for a specified number of cycles. Because there is no standard number of cycles, each table of fatigue strengths must specify the number of cycles for which the strengths are reported. The fatigue strength of nonferrous metals at 100million (108) or 500 million (5 X 108) cycles is erroneously called the fatigue limit. Low-Cycle Fatigue. For the low-cycle fatigue region (N< 104 cycles) tests are conducted with controlled cycles of elastic plus plastic strain,
3
Introduction rather than with controlled load or stress cycles. Under controlled strain testing, fatigue life behavior is represented by a log-log plot of the total strain range, dE, versus the number of cycles to failure (Fig. 3). The total strain range is separated into elastic and plastic components. For many metals and alloys, the elastic strain range, dE eo is equal to the stress range divided by the modulus of elasticity. The plastic strain range, dE p' is the difference between the total strain range and the elastic strain range. Stress-Concentration Factor. Stress is concentrated in a metal by structural discontinuities, such as notches, holes, or scratches, which act as stress raisers. The stress-concentration factor, K" is the ratio of the area test stress in the region of the notch (or other stress concentrators) to the corresponding nominal stress. For determination of K" the greatest stress in the region of the notch is calculated from the theory of elasticity, or equivalent values are derived experimentally. The fatigue notch factor, Kf> is the ratio of the fatigue strength of a smooth (unnotched) specimen to the fatigue strength of a notched specimen at the same number of cycles. Fatigue notch sensitivity, q, for a material is determined by comparing the fatigue notch factor, K J, and the stress-concentration factor, K" for a specimen of a given size containing a stress concentrator of a given shape and size. A common definition of fatigue notch sensitivity is q = (KJ - l)f(K, - 1), in which q may vary between zero (where K J = 1) and 1 (where KJ = K,). This value may be stated as percentage.
Fatigue Crack Propagation In large structural components, the existence of a crack does not necessarily imply imminent failure of the part. Significant structural life may remain in the cyclic growth of the crack to a size at which a critical failure occurs. 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. The growth or extension of a fatigue crack under cyclic loading is principally controlled by maximum load and stress ratio. However, as in crack initiation, there are a number of additional factors that may exert a strong influence, including environment, frequency, temperature, and grain direction. Fatigue crack propagation testing usually involves constant-load-amplitude cy-
10- 3
L..-_ _--I.
10- 1
..J
-l..
--L.
103
10
105
Cycles to failure Fig.3 Typical plot of strain range versus cycles-to-failure for low-cycle fatigue
cling of notched specimens that have been precracked in fatigue. Crack length is measured as a function of elapsed cycles, and these data are subjected to numerical analysis to establish the rate of crack growth, da l d N, Crack growth rates are expressed as a function of the crack tip stress-intensity factor range, dK. The stress-intensity factor is calculated from expressions based on linear elastic stress analysis and is a function of crack size, load range, and cracked specimen geometry. Fatigue crack growth data are typically presented in a log-log plot of da/dNversus s« (Fig. 4). :J.K. ksivTn.
o
5.
10
15
20
25
30 35
1.-----r-"T"T"--r--r-~--,.,..._,_-,-,
10- 1 I---+_--+----;:--et-'~-__l
10- 3 Q)
U
10- 2
>-
~
c:: 10- 4 <:"tl
10- 3
~ 10 -5
10-4
I-~=---+_--+----+---__l
o~ 10- 5
L..-_ _.L-_ _..I.--- _ _-L._ _- '
o
10
20
30
10- 6
40
J.K. MPa\ m Fig. 4 Fatigue crack propagation rate data in 7075-T6 aluminum alloy (R <0)
LIVE GRAPH Click here to view
4
Fatigue Testing
Fatigue Crack Initiation Crack initiation tests are procedures in which a specimen or part is subjected to cyclic loading to failure. A large portion of the total number of cycles in these tests is spent initiating the crack. Although crack initiation tests conducted on small specimens do not precisely establish the fatigue life of a large part, such tests do provide data on the intrinsic fatigue crack initiation behavior of a metal or alloy. As a result, such data can be utilized to develop criteria to prevent fatigue failures in engineering design. 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, complex welded, riveted, or bolted structures, and automotive and aerospace components.
Fatigue Cracking· Fatigue cracks normally result from cyclic stresses that are below the yield strength of the metal. In low-cycle fatigue, however, the cyclic stress may be above the static yield strength, especially in a material with an appreciable workhardening rate. Generally, a fatigue crack is initiated at a highly stressed region of a component subjected to cyclic loading of sufficient magnitude. The crack then propagates in progressive cyclic growth through the cross section of the part until the maximum load cannot be carried, and complete fracture results. Crack Nucleation. A variety of crystallographic features have been observed to nucleate fatigue cracks. In pure metals, tubular holes that develop in persistent slip bands, slip band extrusion-intrusion pairs at free surfaces, and twin boundaries are common nucleation sites. Grain boundaries in polycrystalline metals, even in the absence of inherent grain boundary weakness, are crack nucleation sites. At high strain rates, this appears to be the preferred site. Nucleation at grain boundaries appears to be a geometrical effect, whereas nucleation at twin boundaries is associated with active slip on crystallographic planes immediately adjacent and parallel to the twin boundary. The foregoing processes also occur in alloys and heterogeneous materials. However, alloying and commercial production practices introduce segregation, inclusions, second-phase particles, and other features that disturb the structure. All
of these phenomena have a significant influence on the crack nucleation process. In general, alloying that (1) enhances cross slip, (2) enhances twinning, or (3) increases the rate of work hardening will stimulate crack nucleation. On the other hand, alloying usually raises the flow stress of a metal, thus offsetting its potentially detrimental effect on fatigue crack nucleation. Crack Initiation. Fatigue cracks initiate at points of maximum local stress and minimum local strength. The local stress pattern is determined by the shape of the part and by the type and magnitude of the loading. In addition to the geometric features of a part, features such as surface and metallurgical imperfections can act to concentrate stress locally. Surface imperfections such as scratches, dents, burrs, cuts, and other manufacturing flaws are the most obvious sites at which fatigue cracks initiate. Except for instances where internal defects or special surfacehardening treatments are involved, fatigue cracks initiate at the surface. Relation to Environment. 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. Because it is a synergistic effect of fatigue and corrosion, corrosion fatigue can produce a far greater degradation in strength than either effect acting alone or by superposition of the singular effects. An unlimited number of gaseous and liquid mediums may affect fatigue crack initiation in a given material. Fretting corrosion, which occurs from relative motion between joints, may also accelerate fatigue crack initiation.
Fatigue Testing Regimes The magnitude of the nominal stress on a cyclically loaded component frequently is measured by the amount of overstress-that is, the amount by which the nominal stress exceeds the fatigue limit or the long-life fatigue strength of the material used in the component. The number of load cycles that a component under low overstress can endure is high; thus, the term highcycle fatigue is often applied. As the magnitude of the nominal stress increases, initiation of multiple cracks is more likely. Also, spacing between fatigue striations, which indicate the progressive growth of the crack front, is increased, and the region of final fast fracture is increased in size.
Fatigue Crack Initiation Low-cycle fatigue is the regime characterized by high overstress. The arbitrary, but commonly accepted, dividing line between high-cycle and low-cycle fatigue is considered to be about 104 to 105 cycles. In practice, 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), which in turn depends on the properties of the metal as well as the magnitude of the nominal stress. Presentation of Fatigue Data. High-cycle fatigue data are presented graphically as stress (S) versus cycles-to-failure (N) in S-N diagrams or S- N curves. These are described in the Introduction to this Section along with the symbols and nomenclature commonly applied in fatigue testing. Because the stress in high-cycle fatigue tests is usually within the elastic range, the calculation of stress amplitude, stress range, or maximum stress on the S-axis is made using simple equations from mechanics of materials; i.e., stress calculated using the specimen dimensions and the controlled load or deflection applied axially, in flexure, or in torsion. Figure 5 illustrates a stress-strain loop under controlled constant-strain cycling in a low-cycle fatigue test. During initial loading, the stressstrain curve is O-A-B. Upon unloading, yielding begins in compression at a lower stress C due to
the Bauschinger effect. In reloading in tension, a hysteresis loop develops. The dimensions of this loop are described by its width df (the total strain range) and its height da (the stress range). 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. When the level of cyclic strain is small, the hysteresis loop becomes very narrow. For tests conducted under constant df, the stress range da usually changes with an increasing number of cycles. The common method of presenting low-cycle fatigue data is to plot either the plastic strain range, df p' or the total strain range, df, versus N. When plotted using log-log coordinates, a straight line can befit to the dfp-Nplot. 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, the average value being Y2. This power-law relationship between dfpand Nis known as the CoffinManson relationship. Figure 6 is an example of the typical presentation of low-cycle fatigue test results.
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, however, may be universaltype machines that are capable of conducting all of the above modes ofloading, depending on the fixturing used.
Fig. 5 Stress-strain loop for constantstrain cycling
Fatigue Testing Machine Components Whether simple or complex, all fatigue testing machines consist of the same basic components: a load train, controllers, and monitors. The load train consists of the load frame, gripping devices, test specimen, and drive (loading) system. Typical load train components in an electrohydraulic axial fatigue machine are shown in Fig. 7. The load frame is the structure of the machine that reacts to the forces applied to the specimen by the drive system.
5
6
Fatigue Testing
Ia-... IU
r--..... ......
Q.
'" "1
01
C
10- 1
r-, <,
n
........ --, ~
~ C
~
s
1Il
.~
1Il
a:::
10-
~
"- ..........
2
~
~ ~ O.n
00
~ r-,
10
0
.......
"
Cycles to failure, N Fig. 6
Low-cycle fatigue curve (t1€p versus N) for type 347 stainless steel
Fig. 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. The simplest systems use electric motors to act on test specimens via cams, levers, or rotating grips. In electrohydraulic machines, the motors drive hydraulic pumps to provide service pressure for control of the motion and force of a hydraulic piston actuator. Electromagnetic excitation can be used to excite a mass or inertia system to load a specimen.
Control Systems. The controls and controllers manually or automatically initiate power and test, adjust, and maintain the controlled test parameter(s). Controllers also terminate the test at a predefined status (failure,' load drop, extension, or deflection limit). The control of timevarying deflection or displacement can be obtained in mechanical systems by cam-operated deflection levels, a rotating eccentric mass, or hydraulically through a piston limited by stops. Control in most simple machines and drive systems is obtained via the open-loop mode. In such systems, the magnitude of force and displacement initially set by the control system remains constant throughout the test. Sensors are required to measure the load, strain, displacement, deflection, and cycle count. Some devices provide an output signal to the controller, or to a readout device in the case of uncontrolled parameters. Common sensors are load cells (resistance strain gage bridges calibrated to load) inserted in the load train. Pressure transducers are used in hydraulic or pneumatic actuator devices. Loading fixtures to alter the mode of loading provide versatility. Fixtures can be designed to convert the axial force provided by a hydraulic actuator to perform four-point bending or torsion testing. Similarly, fixtures attached to an oscillating platen of a rotating-eccentric-masstype machine can facilitate axial, bending, and torsion fatigue testing of specimens.
7
Fatigue Crack Initiation
{a}
{e}
{bl
{dl
lei
Ifl
(g)
(a) Standard grip body for wedge-type grips. (b) V-grips for rounds for use in standard grip body. (c) Flat grips for specimens for use in standard grip body. (d) Universal open-front holders. (e) Adapters for special samples (screws, bolts, studs, etc.) for use with universal open-front holders. (f) Holders for threaded samples. (g) Snubber-type wire grips for flexible wire or cable.
Fig. 8
Grip designs used for axial fatigue testing
Grips. Proper gripping is not simply the attachment of the test specimen in the load train. Grip failure sometimes occurs prior to specimen failure. Frequently, satisfactory gripping evolves after specimen design development. Care must be taken in grip design and specimen installation in the grips to prevent misalignment. The grips shown in Fig. 8 are typical of those used for axial fatigue tests. Axial (Direct-Stress) Fatigue Testing Machines The direct-stress fatigue testing machine subjects a test specimen to a uniform stress or strain through its cross section. For the same cross section, an axial fatigue testing machine must be able to apply a greater force than a static bending machine to achieve the same stress. Electromechanical systems have been developed for axial fatigue studies. Generally, these are open-loop systems, but often have partial closed-loop features to continuously correct mean load. In crank and lever machines, a cyclic load is applied to one end of the test specimen through a deflection-calibrated lever that is driven by a variable-throw crank. The load is transmitted to the specimen through a flexure system, which provides straight-line motion to the specimen. The other end of the specimen is connected to a hydraulic piston that is part of an electrohydraulically controlled load-maintaining system that senses specimen yielding. This system automatically and steplessly restores the preset load through the hydraulic piston. Servohydraulic closed-loop systems offer optimum control, monitoring, and versatility in fatigue testing systems. These can be obtained as
component systems and can be upgraded as required. A hydraulic actuator typically is used to apply the load in axial fatigue testing. Electromagnetic or magnetostrictive excitation is used for axial fatigue testing machine drive systems, particularly when low-load amplitudes and high-cycle fatigue lives are desired in short test durations. The high cyclicfrequency of operation of these types of machines enables testing to long fatigue lives (> 108 cycles) within weeks. Bending Fatigue Machines The most common types of fatigue machines are small bending fatigue machines, In general, these simple, inexpensive systems allow laboratories to conduct extensive test programs with a low equipment investment. Cantilever beam machines, in which the test specimen has a tapered width, thickness, or diameter, result in a portion ofthe test area having uniform stress with smaller load requirements than required for uniform bending or axial fatigue of the same section size. Rotating Beam Machines. Typical rotating beam machine types are shown in Fig. 9. The R. R. Moore-type machines (Fig. 9a) can operate up to 10000 rpm. In all bending-type tests, only the material near the surface is subjected to the maximum stress; therefore, in a small-diameter specimen, only a very small volume of material is under test. Torsional Fatigue Testing Machines Torsional fatigue tests can be performed on axial-type machines using the proper fixtures if the maximum twist required is small. Specially
8
Fatigue Testing
A
~Load (b)
(a)
(a) Four-point loading R.R. Moore testing machine. (b) Single-end rotating cantilever testing machine.
Fig. 9 Schematic of rotating beam fatigue testing machines Program
- - - - - Torque feedback - -
---I I I
L----
Angular display feedback
Hydraulic power supply
I
I I
...-----i----r--....,
I I
Specimen
Hydraulic service manifold
Displacement transducer
.,.,C:::}:;;:;;;:;;;:;:l:::;~~,.
Rota ry actu ato r
Fig. 10 Schematic of a servohydraulic torsional fatigue testing machine
designed torsional fatigue testing machines consist of electromechanical machines, in which linear motion is changed to rotational motion by the use of cranks, and servo hydraulic machines, in which rotary actuators are incorporated in a closed-loop testing system (Fig. 10). Special-Purpose Fatigue Testing Machines To perform fatigue testing of components that are prone to fatigue failure (gears, bearings, wire, etc.), special devices have been used, sometimes as modifications to an existing fatigue machine. Wire testers are a modification of rotating beam machines, in which a length of the test wire is
used as the beam and is deflected (buckled) a known amount and rotated. 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. Rolling contact fatigue of ball and roller bearings under controlled lubrication conditions is a specialized field of fatigue testing. Multiaxial Fatigue Testing Machines Many special fatigue testing machines have been designed to apply two or more modes of loading, in or out of phase, to specimens to de-
9
Fatigue Crack Initiation D
termine the properties of metals under biaxial or triaxial stresses.
( ~=====-t-$
Fatigue Test Specimens A typical fatigue test specimen has three areas: the test section and the two grip ends. The grip ends are designed to transfer load from the test machine grips to the test section and may be identical, particularly for axial fatigue tests. The transition from the grip ends to the test area is designed with large, smoothly blended radii to eliminate any stress concentrations in the transition. The design and type of specimen used depend on the fatigue testing machine used and the objective of the fatigue study. 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; i.e., avoiding very low load amplitudes where sensitivity and response of the system are decreased. Several types of fatigue test specimens are illustrated in Fig. 11.
~R
4.8 mm (3116 in.)
D, selected on basis of ultimate strength of material R, 12.7 mm (0.50 in.) (a)
30 mm (13/16 in.)"]
k
50 ~m_1
-$-O~2m.) Tapered D, 12.7 mm (0.50 in.)
Ib)
I'" 90
mm (3:6
i ~ 19
mm (% in.)
l
~~-®R~
~
'\ I 12 mm (0.48 in.)
Effect of Stress Concentration Fatigue strength is reduced significantly by the introduction of a stress raiser such as a notch or hole. Because actual machine elements invariably contain stress raisers such as fillets, keyways, screw threads, press fits, and holes, fatigue cracks in structural parts usually initiate at such geometrical irregularities. 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. Stress concentration can also arise from surface roughness and metallurgical stress raisers such as porosity, inclusions, local overheating in grinding, and decarburization. The effect of stress raisers on fatigue is generally studied by testing specimens containing a notch, usually a V-notch or a U-notch. 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, (2) a stress gradient is set up from the root of the notch toward the center of the specimen, and (3) a triaxial state of stress is produced at the notch root. 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-
:cD
D, 5 to 10 mm (0.20 to 0.40 in.) selected on basis of ultimate strength of material R, 90 to 250 mm (3.5 to 10 in.)
[c]
.25 mm (1.0 in.) D
38 mm (1V2 in.)
~---~ ~3-~ ~ .
5"43'
38 mm (1'12 in.) (d)
D, selected on basis of ultimate strength of material
R, 75 to 250 mm (3 to 10 in.) Ie) (a)Torsional specimen. (bl Rotating cantilever beam specimen. (c) Rotating beam specimen. (d) Plate specimen for cantilever reverse bending. Ie) Axial loading specimen.
Fig. 11
Typical fatigue test specimens
10
Fatigue Testing cent ration factor, K,(see the Introduction to this Section). In some situations, values of K,can be calculated using the theory of elasticity, or can be measured using photoelastic plastic models. The effect of notches on fatigue strength is determined by comparing the S-N curves of notched and unnotched specimens. The data for notched specimens usually are plotted in terms of nominal stress based on the net cross section of the specimen. The effectiveness of the notch in decreasing the fatigue limit is expressed by the fatigue-notch factor, K p This factor is the ratio of the fatigue limit of unnotched specimens to the fatigue limit of notched specimens. For materials that do not exhibit a fatigue limit, the fatigue-notch factor is based on the fatigue strength at a specified number of cycles. Values of KJhave been found to vary with (1) severity of the notch, (2) type of notch, (3) material, (4) type of loading, and (5) stress level.
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. In most cases, a size effect exists; i.e., the fatigue strength of large members is lower than that of small specimens. Precise determination of this phenomenon is difficult. 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. The problems in fatigue testing of large specimens are considerable, and few fatigue machines can accommodate specimens with a wide range of cross sections. Changing the size of a fatigue specimen usually results in variations oftwo factors. First, increasing the diameter increases the volume or surface area of the specimen. The change in amount of surface is significant, because fatigue failures usually initiate at the surface. Secondly, for plain or notched specimens loaded in bending or torsion, an increase in diameter usually decreases the stress gradient across the diameter and increases the volume of material that is highly stressed. Experimental data on the size effect in fatigue typically show that the fatigue limit decreases with increasing specimen diameter. 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.
Table 1 Effect of specimen size on the fatigue limit of normalized plain carbon steel in reversed bending Specimen diameter mm in.
Fatigue limit MPa ksi
7.6 38 152
248 200 144
0.30 l.50 6.00
36
29 21
Surface Effects and Fatigue Generally, fatigue properties are very sensitive to surface conditions. Except in special cases where internal defects or case hardening is involved, all fatigue cracks initiate at the surface. Factors that affect the surface of a fatigue specimen can be divided into three categories: (1) surface roughness or stress raisers at the surface, (2) changes in the properties of the surface metal, and (3) changes in the residual stress condition of the surface. Additionally, the surface may be subjected to oxidation and corrosion. Surface Roughness. In general, fatigue life increases as the magnitude of surface roughness decreases. Decreasing surface roughness minimizes local stress raisers. Therefore, special attention must be given to the surface preparation of fatigue test specimens. Typically, a metallographic finish, free of machining grooves and grinding scratches, is necessary. Figure 12 illustrates the effects that various surface conditions have on the fatigue properties of steel.
Effect of Mean Stress A series of fatigue tests can be conducted at various mean stresses, and the results can be plotted as a series of S-N curves. A description of applied stresses and S-N curves can be found in the Introduction to this Section. For design purposes, it is more useful to know how the mean stress affects the permissible alternating stress amplitude for a given life (number of cycles). This usually is accomplished by plotting the allowable stress amplitude for a specific number of cycles as a function of the associated mean stress. At zero mean stress, the allowable stress amplitude is the effective fatigue limit for a specified number of cycles. As the mean stress increases, the permissible amplitudes steadily decrease. At a mean stress equal to the ultimate tensile strength of the material, the permissible amplitude is zero. The two straight lines and the curve shown in
11
Fatigue Crack Initiation 1000 ,...---,....----,----r---,-------r-----r------,-----,...-----, 800 900 ro 700 a..
~ ~ ~
600
r~~;;::~~;;;;t;;~==~===~~=~===t===~==~
100 'iii
~--_t_---~2.......,_1_--~~=_---=""""
90 -"" vi
.......= _ - _ + - - _ 1 _ - - - - _ t _ - _ _ l
500
W ~ 70
~
~
~ ~ >
60
400
~
Q)
50
~
> ~
>
~ 300 t-----t-----t---+----+----+--"""""~--~""""""----+--___I
40
~ ~
Qi Qj ~
E
E
0
o
u
U; ~
200 I----t----+---+---+-----+---+---+--~.,___f-----~
2
5
10
20
50
100
200
30
u
500
Life, 1000 cycles Fig. 12
Effect of surface conditions on the fatigue properties of steel (302 to 321 HB)
Fig. 13 represent the three most widely used empirical relationships for describing the effect of mean stress on fatigue strength. The straight line joining the alternating fatigue strength to the tensile strength is the modified Goodman law. Goodman's original law included the assumption that the fatigue limit was equal to one third of the tensile strength; this has since been generalized to the relation shown in Fig. 13, using the fatigue strength as determined experimentally. Stress Amplitude. Because stress amplitude varies widely under actual loading conditions, it is necessary to predict fatigue life under various stress amplitudes. The most widely used method of estimating fatigue under complex loading is provided by the linear damage law. This is a hypothesis first suggested by Palmgren and restated by Miner, and is sometimes known as Miner's rule. The assumption is made that the application of n.cycles at a stress amplitude S;, for which the average number of cycles to failure is N;, causes an amount of fatigue damage that is measured by the cumulative cycles ratio n;/N;, and that failure will occur when "'i.(n;/ N;) = 1. This method is not applicable in all cases, and numerous alternative theories of cumulative linear damage have been suggested. Some considerations of redistribution of stresses have been clarified, but there is as yet no satisfactory approach for all situations.
Fatigue strength, S CI)"
vi ~ ~
OJ C
Gerber's parabola
/
/
Modified Goodman line Tensile strength, Su
Mean stress, Sm As shown by the modified Goodman line. Gerber's parabola. and Soderberg line. See text for discussion.
Fig. 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. The test is then continued to fracture at a different amplitude. Alternatively, the stress can be changed from one stress amplitude to another at regular intervals; such tests are known as block, or interval, tests. These tests do not simulate service conditions, but may serve a useful purpose for assessing the linear damage law and indicating its limitations.
12
Fatigue Testing Corrosion Fatigue
Stress-intensity factor range UK). ksi\.
20
10
Corrosion fatigue is the combined action of repeated or fluctuating stress and a corrosive environment to produce progressive cracking. Usually, environmental effects are deleterious to fatigue life, producing cracks in fewer cycles than would be required in a more inert environment. Once fatigue cracks have formed, the corrosive aspect also may accelerate the rate of crack growth. In corrosion fatigue, the magnitude of cyclic stress and the number of times it is applied are not the only critical loading parameters. Timedependent environmental effects also are of prime importance. When failure occurs by corrosion fatigue, stress-cycle frequency, stresswave shape, and stress ratio all affect the cracking processes.
100
,
q I I
a ~
6~
]
.
t----
da dN
~ C(~Kln
I~
dO
: W '0 Ie? if
sx;
I 5
I
6
~~o
I I I
i
'I
~
.9' r;~
i Region 1: slow crack I- growth
I
Fatigue Crack Propagation
fil.
50
I
I
1
Region 3:
rapid unstable crack
growlh
I rI
!
r--
I I 1
Region 2: power-taw behavior
I,c
1;1 I,j> I
Fatigue failure of structural and equipment components due to cyclic loading has long been a major design problem and the subject ofnumerous investigations. Although considerable fatigue data are available; the majority has been concerned with the nominal stress required to cause failure in a given number of cyclesnamely, S-N curves. Usually, such data are obtained by testing smooth or notched specimens. With this type of testing, however, 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. 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.
Fatigue Crack Propagation Test Methods The general nature of fatigue crack propagation using fracture mechanics techniques is summarized in Fig. 14. A logarithmic plot of the crack growth per cycle, da/ dN, versus the stressintensity factor range, I:!..K, corresponding to the load cycle applied to a specimen is illustrated. The da/ dN versus I:!..K plot was constructed of
I
8
20
10
30
40
50 60
Suess-tntensnv teeter range UK), MPa \
80 100
m
Yield strength of 470 MPa (70 ksi). Test conditions: R = 0.10; ambient room air, 24°C (75 OF).
Fig. 14 Fatigue crack growth behavior of ASTM A533 B1 steel
data on five specimens of ASTM A533 HI steel tested at 24 0 C (75 0 F). A plot of similar shape is anticipated with most structural alloys; the absolute values of da/dNand I:!..K, however, are dependent on the material. Results of fatigue crack growth rate tests for nearly all metallic structural materials have shown that the da/ dN versus I:!..K curves have three distinct regions. The behavior in Region I (Fig. 14) exhibits a fatigue crack growth threshold, I:!..K"" which corresponds to the stressintensity factor range below which cracks do not propagate. At intermediate values of I:!..K (Region II in Fig. 14), a straight line usually is obtained on a log-log plot of I:!..K versus da/ dN. This is described by the power-law relationship: da dN
= C(I:!..K)"
where C and n are constants for a given material and stress ratio.
13
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. Data for four martensitic steels fall within a single band, as shown in Fig: 15. The upper bound of scatter can be obtained from: da dN = 0.66 X
1O-8(~K)2.25
where a is given in inches, and ~K is given in ksiyTr;. For some steels, the stress ratio and mean stress have negligible effects on the rate of crack growth in Region II. Also, the frequency of cyclic loading and the waveform (sinusoidal, triangular, square, trapezoidal) do not affect the rate of crack propagation per cycle of load for some steels in benign environments. At high ~Kvalues (Region III in Fig. 14), unstable behavior occurs, resulting in a rapid increase in the crack growth rate just prior to complete failure of the specimens. There are two possible causes of this behavior. First, the increasing crack length during constant load testing causes the peak stress intensity to reach the fracture toughness, K'n of the material, and the unstable behavior is related to the early stages of brittle fracture. Second, the growing crack reduces the uncracked area of the specimen sufficiently for the peak load to cause fully plastic limit load behavior. The first possibility is operative for high-strength, low-toughness metals, 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. The second possibility, plastic limit load behavior, is common for ductile metals, particularly if K/cis high. When plastic limit load behavior causes unstable crack growth, ~K values have no meaning, because the limitations of linear elastic fracture mechanics have been exceeded. Here, the use of the J-integral concept, crack-opening displacement, or some other elastic-plastic fracture mechanics approach is more appropriate than ~K for correlating the data. Standardized testing procedures for measuring fatigue crack growth rates are described in ASTM Standard E 647. This method applies to medium to high crack growth rates-that is, above 10-8 tu] cycle (3.9 X 10-7 ui.] cycle). Procedures for growth rates below 10-8 in] cycle are
Stress-intensity factor range (.:lK), ksiV'Ti1."
102
10
10- 2
o o • •
Ql
"0
>
~
E E
12 Ni steel 10 Ni steel HY-80 steel HY-130 steel
Ql
10- 4 "0 > ~ c::
10- 3
......:
<:
<: ~
~
~
~
10- 5
2:! s:
~
i
~
~
s:
10- 4
~
e
e
Ol
Ol
-" 0
-" 0
10- 6
~ U
co
L.
U
10- 5 t--t-l~tf--+--+--+----I
2 x 10- 6
'---_..I..-_..L-_..I..-_....L_--L--'
2
5
10
20
50
100
200
Stress-intensity factor range (.:lKl. MPaVm Fig. 15 Summary of fatigue crack growth data for martensitic steels
under consideration by ASTM. For applications involving fatigue lives of up to about 106 load cycles, the procedures recommended in ASTM E 647 can be used. Fatigue lives greater than about 106 cycles correspond to growth rates below 10-8 in] cycle, and these require special testing procedures, which are related to the threshold of fatigue crack growth illustrated in Fig. 14. ASTM E 647 describes the use of centercracked specimens and compact specimens (Fig. 16 and 17). The specimen thickness-to-width ratio, B/W, is smaller than the 0.5 value for K/c tests; the maximum B/W values for centercracked and compact specimens are 0.125 and 0.25, respectively. With the thinner specimens, crack length measurements on the sides of the specimens can be used as representations of through-thickness crack growth behavior. For tension-tension fatigue loading, the K; loading fixtures frequently can be used. For this type of loading, both the maximum and minimum loads are tensile, and the load ratio, R = P min/Pmax' is in the range 0 < R < 1. A ratio of
14
Fatigue Testing Two holes W/3 diam
-r W
~----+-----J_l !I ! I
!I I
2a n i.sthe machined notch; a is the crack length; B is the
specimen thickness.
Fig. 16 Standard center-cracked tension specimen for fatigue crack propagation testing when the width (WI of the specimen ';;;;75 mm (3 in.)
R= 0.1 is commonly used for developing data for comparative purposes. Testing often is performed in laboratory air at room temperature; however, any gaseous or liquid environment and temperature of interest may be used to determine the effect of temperature, corrosion, or other chemical reaction on cyclic loading. Data Analysis. For constant-amplitude loading, a set of crack-length versus elapsed-cycle data (a versus N) is generated, with the specimen loading, Pmax and P min' generally held constant. Figure 18illustrates a typical a versus N plot. The minimum crack-length interval, 6.a, between data points (see Fig. 18)should be 0.25 mm (0.01 in.) or ten times the crack-length measurement precision, which is defined as the standard deviation on the mean value of crack length determined for a set of replicate measurements. 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. Crack measurement intervals are recommended in ASTM E 647 according to specimen type. For compact-type specimens: 6.a ~ 0.04 Wfor 0.25 ~!!.- ~ 0.40
W
6.a ~ 0.02 W for 0.40 ~ !!.- ~ 0.60 W 6.a~
a
0.01 Wfor -;;::: 0.60 W
For center-cracked tension specimens: 6.a ~ 0.03 Wfor 2a W 6.a ~ 0.02 W for
< 0.60
~ > 0.60 W
Fatigue crack growth rate data can be calculated by several methods. The most commonly used methods, however, are the secant and incremental polynomial methods. The secant method consists of the slope of the straight line connecting two adjacent data points. This method, although simpler, results in more scatter in measured crack growth rate. The incremental polynomial method fits a secon~-order ,Polynomial expression (parabola) to typically five to seven adjacent data points, and the slope of this expression is the growth rate. The incremental polynomial method eliminates some of the scatter in growth rate that is inherent in fatigue testing. Numerous relationships have been generated to correlate crack growth rate and stress-intensity data. The most widely accepted relationship is that proposed by Paris. This is a linear relationship when plotted on log-log coordinates and generally yields a reasonable fit to the data in Region II (see Fig. 14) of the crack growth regime. Other relationships based on the Paris equation, such as the commonly used Forman equatl?n, are used to represent the variation of da/ dN with other key variables, including load ratio, R and the critical K value, K" at which rapid frac~ ture of the specimen occurs (Region III in Fig. 14). The Forman equation is: da = C(6.K)" dN (1 - R)(K,. - 6.K)
where Cand n are material constants of the same types as those in the Paris equation, but of different values. 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.K, which is not described by the Paris equation. Additionally, the Forman equation describes the frequently observed increase in da/ dN asso-
15
Fatigue Crack Propagation Two holes
0.25Wdiam
r
t t
0.6W
0.275W
1
t t
0.275W
r
0.6W
~~_1_ 1 + - - - a ----;~ ( - + - - - - - - w-------;~I ) 1
Allowable thickness: W/20 s B s W/4 Minimum dimensions: W = 25 mm (1.0 in.)
an
= 0.20W
~------1.25W -------:;~I
Fig. 17 Standard compact-type specimen for fatigue crack propagation testing (see Fig. 16 for explanation of symbols)
ciated with an increase in R from 0 toward I. 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. When only ~Kin Region II is involved, the less complex Paris equation may be used.
Cyclic Crack Growth Rate Testing in the Threshold Regime
55
2.2 (
50
E E 45
.r=
I
40
~ 35 u
1.6
(J
00
20 125
v
n
Oo c
150
1.0 175
200
225 250
.><
1.2
30 25
C
s: 0, 1.4 Q)c
'rP0a- J
.><
~
1.8
",,0
(J
Cyclic crack growth rate testing in the lowgrowth regime (Region I in Fig. 14) complicates acquisition of valid and consistent data, because the crack growth behavior becomes more sensitive to the material, environment, and testing procedures under this regime. Within this regime, the fatigue mechanisms of the material that slow the crack growth rates are more significant. The precise definition of the cyclic crack growth rate threshold, ~K'h' varies significantly.
~
i~'
OJ
2.0
275
300
3
Cycles. 10 N Fig. 18 Crack growth versus constantamplitude stress cycles for a Fe-10Ni8Co-1 Mo high-strength steel
...
ro U
16
Fatigue Testing The most accurate definition would be the stressintensity value below which fatigue crack growth will not occur. It is extremely expensive to obtain a true definition of IJ.K,h, and in some materials a true threshold may be nonexistent. Generally, designers are more interested in the near-threshold regime, such as the IJ.Kthat corres~onds to a fatigue crack growth rate of 10-8 to 10- 0 tn] cycle (3.9 X 10-7 to 10-9 in.j cycle). Because the duration of the tests increases greatly for each additional decade of near-threshold data (10- 8 to 10-9 to 10-10, etc., m/cycle), the precise design requirements should be determined in advance of the test.
Short cracks that join long crack behavior
q/;/"
Behavior of Short Cracks Recently, 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. A short crack is difficult to define. It may be small compared to the microstructure of the material to be studied (I to 50 /.Lm) when the concepts of continuum mechanics are of interest. It can also be small compared to the plastic zone size (10 to 1000 /.Lm). In this situation, linear elastic fracture mechanics might be replaced with elastic-plastic fracture mechanics. The crack may also be physically small (500 to 1000 /.Lm) when crack closure, crack tip shape, environment, and growth mechanisms are of concern. Figure 19 schematically illustrates the possible behavior of short cracks.
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. The two most widely used types of specimens are the centercracked tension specimen and the compact-type specimen (see Fig. 16 and 17). However, any specimen configuration with a known stressintensity factor solution can be used in fatigue crack growth testing, assuming that the appropriate equipment is available for controlling the test and measuring the crack dimensions. Stressintensity factor solutions for center-cracked tension and compact-type specimens are given in Table 2. Consideration of the range of application of the stress-intensity solution of a specimen configuration is very important. Many stress-inten-
//
\',
\
/
/
/
\
Short cracks that behave as long cracks
./
\---
\
\
Short cracks that become non propagating cracks
Stress-intensity factor range Fig. 19 Typical short crack behavior
sity expressions are valid only over a range of the ratio of crack length to specimen width (a/W). For example, the expression given in Table 2 for the compact-type specimen is valid for a/ W > 0.2; the expression for the center-cracked tension specimen is valid for 2a/W< 0.95. The use of stress-intensity expressions outside their applicable crack-length region can produce significant errors in data. The size of the specimen must also be appropriate. To follow the rules of linear elastic fracture mechanics, the specimen must be predominantly elastic. However, unlike the requirements for plane-strain fracture toughness testing, the stresses at the crack tip do not have to be maintained in a plane-strain state. The stress state is considered to be a controlled test variable. The material characteristics, specimen size, crack length, and applied load will dictate whether the specimen is predominantly elastic. Because the loading mode of different specimens varies significantly, each specimen geometry must be considered separately. Notch Preparation. The method by which a notch is machined depends on the specimen
Fatigue Crack Propagation Table 2 Stress-intensity factor solutions for standardized (ASTM E 647) fatigue crack growth specimen geometries
Center-cracked tension specimens (Fig. 16)
sr I:!.K=
Ii V
~
2W
sec
tra
2
2a; expression . va1'" 2a 0.95 where a = id 10r-<
W
W
Compact-type specimens (Fig. 17) _ I:!.K -
I:!.P(2
+ a)
ru; 3/2 (0.886 + 4.64a - 13.32a By W(l-a)
2
+ 14.Tl« 3 -
4
5.6a )
. va liId f or-» a 0.2 where a = -a ; expression
W
W
material and the desired notch root radius (p). Sawcutting is the easiest method, but is generally acceptable only for aluminum alloys. For a notch root radius of p ~ 0.25 mm (0.010 in.) in aluminum alloys, milling or broaching is required. A similar notch root radius in low- and medium-strength steels can be produced by grinding. For high-strength steel alloys, nickelbase superalloys, and titanium alloys, electrical discharge machining may be necessary to produce a notch root radius of p ~ 0.25 mm (0.010 in.). 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. To decrease the amount of time needed for precracking to occur, 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. Load generally is reduced uniformly to avoid transient effects. 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. Therefore, the step size of the load during precracking should be minimized. Reduction in the maximum load should not be greater than 20% of the previous load condition. As the crack approaches the final desired size, this percentage may be decreased. Gripping of the specimen must be done in a manner that does not violate the stress-intensity solution requirements. For example, in a singleedge notched specimen, it is possible to produce a grip that permits rotation in the loading of the specimen, or it is possible to produce a rigid grip.
Each of these requires a different stress-intensity solution. In grips that are permitted to rotate, such as the compact-type specimen grip, the pin and the hole clearances must be designed to minimize friction. It is also advisable to consider lateral movement above and below the grips. Gripping arrangements for compact-type and center-cracked tension specimens are described in ASTM E 647. For a center-cracked tension specimen less than 75 mm (3 in.) in width, a single pin grip is generally suitable. Wider specimens generally require additional pins, friction gripping, or some other method to provide sufficient strength in the specimen and grip to prohibit failure at undesirable locations, such as in the grips. Grips designed for compact-type specimens are illustrated in Fig. 20.
Crack-Length Measurement Techniques Several different techniques have been developed to monitor the initiation, growth, and instability of cracks, including optical (visual and photographic), electrical (eddy current and resistance), compliance, ultrasonic, and acoustic emission monitoring techniques. Optical Crack Measurement Techniques Monitoring of fatigue crack length as a function of cycles is most commonly conducted visually by observing the crack at the specimen surfaces with a traveling low-power microscope at a magnification of 20 to 50X. Crack-length measurements are made at intervals such that a nearly even distribution of da] dN versus ~K is achieved. The minimum amount of extension be-
17
18
Fatigue Testing p
t (a)
~--al~ p
Fig. 20 Grips designed for fatigue crack propagation testing of compact-type specimens (courtesy of MTS Systems Corp.)
tween readings is commonly about 0.25 mm (0.10 in.). The optical technique is straightforward and, if the specimen is carefully polished and does not oxidize during the test, produces accurate results. However, the process is time consuming, subjective, and can be automated only with complicated and expensive video-digitizing equipment. In addition, many fatigue crack growth rate tests are conducted in simulatedservice environments that obscure direct observation of the crack. 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, 0, and the tensile load, P, with the displacement measured along, or parallel to, the load line. Fig-
p
(b) (a) C(aD)
= 0D/P, (b) C(a1) = 01 /P
Fig. 21 Schematic of the relationship between compliance and crack length
ure 21 illustrates that the more deeply a specimen is cracked, the greater the amount of 0 measured for a specific value of tensile load. Compliance can also be defined for shear and torsional loads applied to cracked specimens, and crack extension under these loading modes can be similarly determined. Specimen load is simultaneously measured by an electronic load cell and conditioner / amplifier system, and the output is directed to the same data-acquisition system. A generalized schematic of the circuits involved is shown in Fig. 22.
19
Fatigue Crack Propagation
Displacement gage
Load
a, a 1 a2
-+----... )
....
'Lr
x-v recorder
t
Specimen
a
Gage condition
/P'~
J=1
Load cell
I ± 10 V dc
Load cell condition ± 10 V dc
Fig. 22
Components of a compliance measurement system
The required sensitivity of the systems depends on specimen geometry and size; in general, noise-free, amplified output on the order of I V dc per I mm (0.04 in.) of deflection is satisfactory. Similarly, for the load range applied to the specimen, an approximately I V de change in signal from the load cell is required for accurate calculation of the compliance. Electric Potential Crack Monitoring Technique The electrical potential, or potential drop, technique has gained increasingly wide acceptance in fracture research as one of the most accurate and efficient methods for monitoring the initiation and propagation of cracks. This method relies on the fact that there will be a disturbance in the electrical potential field about any discontinuity in a current-carrying body, the magnitude of the disturbance depending directly on the size and shape of the discontinuity. For the application of crack growth monitoring, the electric potential method entails passing a constant current (maintained constant by external means) through a cracked test specimen and measuring the change in electrical potential across the crack as it propagates. With increasing crack length, the uncracked cross-sectional area of the test piece decreases, its electrical resistance increases, and thus the potential difference between two points spanning the crack rises. By monitoring this potential increase, Va' and comparing it with some reference potential, Vo , the
crack length to width ratio, a/ W, can be determined through the use of the relevant calibration curve for the particular test piece geometry concerned, Crack Growth Studies. By far the most useful application of the electrical potential method has been in measurements of crack length during crack propagation, where it has been utilized to monitor almost all mechanisms of subcritical crack growth and most notably to follow fatigue crack growth. Typical crack propagation rates derived from direct current potential measurements are shown in Fig. 23 for tests on a 2.25CrIMo steel in air, gaseous hydrogen, and hydrogen sulfide environments.
Electromechanical Fatigue Testing Systems The primary function of electromechanical fatigue testers is to apply millions of cycles to a test piece at oscillating loads up to 220kN (50000 lbf) to investigate fatigue life, or the number of cycles to failure under controlled cyclic loading conditions. Variables associated with fatigue-life tests are frequency of loading and unloading, amplitude of loading (maximum loads and minimum loads), and control capabilities. The fundamental data output requirement is the number of cycles to failure, as defined by the application, A variety of electromechanical fatigue testers have been developed for different applications. Forced-displacement, forced-vibration, rota-
20
Fatigue Testing
Table 3
Comparison of electromechanical fatigue systems
Parameter
Forced displacement
Tension Compression Reverse stress Bending Frequency range Load range
Forced vibration
. Yes Yes . . Yes . Yes . Fixed . Typically < 450 N « 1001bf)
Type: Control Mode
Open-loop Displacement
. .
Maximum deflection Advantages
Yes Yes Yes Yes Fixed, 1800 rpm Up to 220 kN (500001bf) Open-loop Load
.
25.4 mm (l.00 in. Versatile, efficient, durable Fixed frequency, Iii ited control (open loop)
Simple, straightforward No load control, very limited applications (soft samples)
Disadvantages
Alternating stress intensity (:>K), ksi vln, 678910
I
E
• 0
• •'" 0
l>
I
I
I
Environment Moist air Dry hydrogen Air Dry hydrogen Hydrogen sulfide Molstair Dry hydrogen
~
~
::
I~rt
t
V~
r
I 'I II ,
I
20
Frequency. Hz 50 50 5 5 5
R
0.05 0.05 0.1 0.1 0.1 0.75 0.75
I
40
~#
/
nO
# 00 '"
~
• fill'
~
t4,f-
~ lJ
.. ..
60 10 80 I
.
-
10" ,
.
~
U
10-.
~
. ..
1)
~
-
10- 1
~
s:
~
'U
One lattice -.. spacing -- 10-. per cycle
I
-
0 0
> u
.~
- 10'
oOd9
~.'l>~ ~ rP
I
rPJ pO
od
!II
50
~I
I"'" '"
i-
,
30
I
I
10
ern
..
~
u
U
•
Threshold :>K",
t 6
7
8 9 10
20
30
40
50 60 10 80 90
Alternating stress intensity (:>K), MPa\m Data derived from direct current potential measurements in martensitic 2.25Cr-1 Mo steel (SA542-C12) at R =0.05 to 0.75 in air, hydrogen. and hydrogen sulfide at ambient temperature.
Fig. 23
Fatigue crack propagation data over a wide spectrum of growth rates
tional bending, resonance, and servomechanical systems are discussed in this article and are compared in Table 3. Other specialized electromechanical systems are available to perform specific tasks. Forced-Displacement Systems Forced-displacement motor-driven systems are the simplest type of electromechanical fatigue testers. They effectively reproduce service
environments that impart fixed, reciprocating displacements to a component or test piece. An electric motor-driven flywheel is used to carry a loading arm at a variable distance from the center of rotation, much in the same manner as a connecting rod in an automotive engine. This rotational displacement is transformed into a guided, vertical displacement and is used to fatigue the specimen. Although load can be monitored in such sys-
21
Fatigue Crack Propagation
Rotational bending
Resonance
Servomechanical
No No Yes Yes 0-10000 rpm
Yes Yes Yes Yes 40-300 Hz Up to 180 kN (400001bf)
Yes Yes Yes Yes 0-1 Hz Up to 90 kN (200001bf)
Open-loop Rotation/ bending
Closed-loop Load
fficient, durable, simple .otational bending only, limited applications
l.0 mm (0.040 in.) Fully closed-loop, extremely efficient Operating frequency directly proportional to sample stiffness
Closed-loop Load, displacement, strain 100 mm (4 in.) Fully closed-loop, high precision Low frequency only
tems, the fixed displacement precludes the ability to control load, which is a function of specimen characteristics. Therefore, the load generally drops as failure progresses. These systems typically are custom-built, inexpensive fatigue machines, used primarily for bend tests on soft samples in which load control, high frequencies, and large loads are not required. Forced-Vibration Systems Forced-vibration motor-driven systems were the first production fatigue testers in commercial use. The centrifugal forces of an imbalanced rotor is used to impart a cyclic load to the test piece. In operation, an electric motor is used to rotate an eccentric mass via flexible couplings. The rotating mass is mounted in a frame that is guided by flexure plates to restrict movement to vertical motion only. The centrifugal force produced by the rotating eccentric mass (m) is transmitted through the vertically guided frame to the test piece. The horizontal component of the centrifugal force is absorbed by the restraining flexure plates. Because the centrifugal force usually is totally absorbed by the mounting frame (of mass M), the inertial reaction is separated from the centrifugal force in such a way as to transmit only the centrifugal forces to the specimen. This technique involves the use of frame-support compensator springs; the natural frequency of the spring (K)/mass (M) system is tuned to the revolutions per minute of the motor. Thus, neither the specimen nor the rotating eccentric mass (m) "sees"
an inertial reaction from the frame, because the inertial effects of the frame are totally compensated for by the frame support springs (not the specimen). This technique has two requirements: the rotating frequency (w) must be kept constant and the mass of mounting frame (M) must be kept constant. Consequently, the loading frequency of the device is fixed at 1800 rpm, and masses must be added or removed from the frame to compensate for fixturing to keep M constant. The magnitude of the dynamic load is determined by placing the rotating mass at a known distance from the axis ofrotation (r). Because w, m, M, and K are known, the force on the specimen, F, is calibrated directly as a function of r as follows: F
= Mw 2r (centrifugal) -
Ma, (inertial)
+ Kz (spring compensated)
where a, is the acceleration of the frame in the z direction, and Kz is the spring-compensated displacement in the z direction. Because Ma, is tuned to equal Kz, F = Mw 2r. Thus, the forced-vibration rotating eccentric mass system is an open-loop, load-controlled system with the ability to accommodate up to 25 mm (1.0 in.) of total sample deflection at loads up to 220 kN (50000 lbf) using special fixtures. The mean or static load, onto which the dynamic load is superimposed, is achieved by preloading the inertia compensator spring, K. Through special fixturing, forced-vibration devices are capable of testing in tension, com-
22
Fatigue Testing pression, bending, torsion, or reverse st~ess. Although servo-controlled, mean-load-maintenance systems are available, the open-loop nature of the system prevents direct load measurement or control, which is characteristic of closed-loop systems. The load applied to the specimen is assumed to be a function of r, and a graduated scale is provided to permit reasonably accurate setup. Rotational Bending Systems Rotational bending systems effectively apply reversed loading to the outer surface of rods or shafts. The basic operating principle ofthe rotating beam consists of the use of a motor to rotate a shaft of known dimensions around its longitudinal axis. By applying a known static force at the end of the shaft, a bending moment can be applied to the test section, the outer surface of which oscillates between tension and compression during each rotation. The cantilevered specimen, however, is subjected to a nonuniform bending moment, which is large at the supported end of the specimen ~nd zero at the free end. To produce a more meanmgful uniform bending moment throughout the tes~ piece, a specially designed tapered specimen should be used or bending moments should be applied to each end of the specimen. Figure 24 illustrates the rotating-beam operating mechanism and the resulting stress distribution in the specimen. Gage area
Drive motor
I Tension '----.. )
Hotation
0~:f:
A
Bending
mom,",
Compression Fig.24 Schematic ofthe rotating-beam operating mechanism and the resulting stress distribution in the specimen
Resonance Systems A high-speed fatigue testing system was developed by Amsler that operated at 40 to 300 Hz, achieved high loads (up to 90 kN, or 20000 lbf), and consumed minimal energy. It is based on a resonant spring/mass system, in which the specimen is used, like a spring, as an integral part of the oscillating mechanism. The fatigue load, in the form of a sine wave, is achieved by preloading the sample in the frame via a complex optomechanical procedure and dynamically loading the sample at the natural oscillating frequency of the spring/ mass system. The preload is maintained automatically during the test. The dynamic load is achieved by pulsing an electromagnet at the natural frequency of the spring/ mass system. During resonance, the electromagnet restores any hysteresis energy lost during the previous cycle, thereby maintaining a constant, controllable dynamic load. Capable of tension, compression, bending, torsional, and reverse-stress fatigue tests, the Amsler resonant fatigue testers were instrumental in obtaining the vast amount of fatigue data currently available. The resonant system is based on a similar principle, but incorporates solid-state technology to achieve fully closed-loop control of mean and dynamic loads. This system uses dual opposing masses (unlike the single oscillating mass/ seismic base of earlier systems), linked by the specimen to achieve vibration-free resonance. A strain gage load cell, in series with the specimen, senses the load and automatically triggers the electromagnet to achieve self-tuning capability. The mean load is achieved by physically moving the upper mass up or down to achieve tension or compression, respectively; the dynamic load is achieved by varying the width of the pulse to the magnet beneath the lower mass. The dynamic load, like the mean load, is electronically ma~n tained at a preset command level through solidstate closed-loop circuitry. The remainder of the controls and mechanisms associated with the resonant fatigue system maintain a preset air gap between the magnet and the oscillating lower mass, maintain preset loading conditions (S~1Ut ting down at preset load levels or frequencies), and power the electromagnet. The high efficiency of resonant systems makes them well suited to high-cycle fatigue tests, in which closed-loop load control, high loads (up to 180 kN, or 40000 lbf), low power consumption (around 750 W maximum for closed-loop sys-
23
Fatigue Crack Propagation Electronic demand signal
1-----,---- c
(bl
(a) Typical components. (b) Transfer functions. See text for details and explanation of symbols.
Fig. 25
Simplified block diagram for a negative-feedback closed-loop testing machine
tems), and high throughput are required. These systems tolerate minimal hysteresis and produce optimum testing results when used with stiff metallic samples. Closed-Loop Servomechanical Systems The most recent development in electromechanical fatigue testers is based on an electric actuator/load frame assembly. The system closely resembles its servohydraulic counterpart in that it consists of an actuator, a load frame, a load cell, a power supply, and a solid-state closedloop electronic control console. Closed-loop systems compare live feedback signals to an input command signal to maintain accurate control of preset conditions. The closed-loop servomechanical system is, by virtue of its design, primarily intended for low-cycle and creep-fatigue studies.
Servohydraulic Fatigue Testing Systems Servohydraulic testing machines are particularly well suited for providing the control capabilities required for fatigue testing. Extreme demands for sensitivity, resolution, stability, and reliability are imposed by fatigue evaluations. Displacements may have to be controlled (often for many days) to within a few microns, and forces can range from 100 kN to just a few newtons. This wide range of performance can be obtained with servomechanisms in general and, in particular, with the modular concept of servohydraulic systems.
Usually, the problem of selecting the appropriate system is simply a matter of optimizing the various components to form a system best suited to the given testing application. In this section, the principles underlying closed-loop servo systems are discussed briefly. In addition, the interaction between system components is illustrated, and a brief description of their operating principles and characteristics is provided. With any type of control system, the objective is to obtain an output that relates as closely as possible to the programmed input. In a fatigue testing system, it may be desired to vary the force on a specimen in a sinusoidal manner, at a frequency of 1 Hz over a force range of 0 to 100 kN (0 to 22000 lbf). The only practical means to accomplish this with precision is through the use of a negative-feedback closed-loop system. An overview of the basic principles of operation of negative-feedback systems is provided in Fig. 25. The blocks shown in Fig. 25(a) represent a group oftypical components of a testing machine. The transfer functions of each of these blocks can be combined to produce the more simplified diagram shown in Fig. 25(b). Placement of the switch, S I' has been added to the diagram to permit analysis of the system when it is open (no feedback, or an open-loop condition) and when it is closed (providing feedback to the system). The equation governing this simplified open-loop system is: C= KoD
where Crepresents the controlled output, K; rep-
24
Fatigue Testing resents the open-loop transfer function, and D represents the electronic demand signal. Therefore, the output is simply proportional to the system demand if K is a constant. Unfortunately, K is seldom a constant, because it can be influen"ced by several common system variations. The electronic components may drift slightly, or the~r gain may vary. The behavior of the hydraulic components may change with tempera~ure, contamination, or wear, and the mechamcal components may vary because of thermal effects or friction. Servohydraulic System Components Many commercially manufactured units are available for each component in a typical servohydraulic testing system. . The programmer supplies the command signal to the system, which is generally an analog of the desired behavior of the controlled parameter. For example, assume the same test conditions as previously discussed (control the force on the specimen in a sinusoidal manner at a frequency of I Hz and a force range of 0 to 100 kN). In this instance, the programmer might be set to produce an electronic signal with a sinusoidal waveform that has a frequency of I Hz and a voltage output of 0 to 10 V. The analog is: 1 V represents 1000 N. The system can then be adjusted to produce the correct output. Any change in the programmer signal will result in a corresponding change in the controlled parameter. The servo-controller makes most of the adjustments necessary to optimize system performance. For example, it compares the command signal with a signal produced by the controlled parameter (stress or strain, for example) and relays a correction signal, if needed, to the control device in the system (usually a flow-control servo-valve). A servo-controller incorporates numerous other compensatory features, such as:
• 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 . • 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
• Auxiliary functions such as recorder signal conditioning, calibration, 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. Within the control loop, it is the intermediary between the low-power servo-controller and the hydraulic ram, which can supply large force~ a~d displacements to the specimen. Characte~ISt1CS of the device are such that the output flow IS approximately proportional to the input current when the output pressure is constant. Also, the output pressure is approximately proportlOn~1 to the square of the input current when the flow IS constant. Hydraulic rams, or actuators or cylinders, furnish the forces and displacements required by the testing system. 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. 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. Under static conditions (very little flow), the maximum force capability of the ram will approach the hydraulic supply pressures multiplied by the effective area. 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. Load Cells. The strain gage load cell is the most widely used force-measuring and feedback device in closed-loop fatigue machines. An external applied force causes the elastic deformation of an internal member to which a strain gage bridge has been attached. An electronic sign~l that is proportional to the resistance change m the bridge and to the applied force can thus .b.e produced. Some load cells are designed specifically for fatigue evaluations. Variable features include sensitivity, natural resonant frequency, temperature stability, fatigue rating, linearity, hysteresis, deflection constant, load capacity, overload rating, resistance to extraneous loading, and compatibility with t?e testing machine and fixtures. Most commercially available cells are very competitive with respect to these features. Load Frames. In a fatigue machine, the reaction forces to the specimen and to the housing of
Fatigue Crack Propagation the ram are supplied by the load frame. Many styles ofload frames are available, but for fatigue purposes the frames should be customized. The requirements of good high-frequency response demand that there be high axial stiffness in the load frame. When a deflection occurs in the load frame, additional flow is required from the servo-valve. Therefore, this deflection should be minimal in comparison with the deflection imparted to the specimen. In addition, because fatigue specimens must be subjected to fully reversed loading (i.e., compressive as well as tensile forces), lateral rigidity
must be increased to resist bending. This is generally considered necessary in the design of fatigue machines. The extra rigidity can be obtained by increasing the diameter of the support columns or by utilizing three- or four-column configurations. Exceptional alignment is required of load frames used in fatigue evaluations to minimize undesirable bending forces. In addition, some means is usually provided to refine the alignment with manual adjustments when necessary. A strain-gaged specimen can be used to make this evaluation.
25
l-1.
S-N Curves Typical
27
for Steel
Schematic S-.V curves for a material at various stress ratios. l!TS and I’S indicate ultimate tensile strength and yield strength, respectiveI), in uniaxial tensile testing.
The results of fatigue tests are usually plotted as maximum stress or stress amplitude to number of cycles. .V. to fracture using a logarithmic scale for the number of cycles. Stress is plotted on either a linear or a logarithmic scale. The resulting tune of data points is called an S-,Vcur\e. A family of S-,Z’curves for a material tested at various stress ratios is shown schematically in the above curves. Stress ratio is the algebraic ratio oftwo specified stress Lalues in a stress cycle. Twocommonl~ used stress ratiosare the ratio. A.ofthealternatingstressamplitudeto the mean stress (A = Sa,‘Sm) and the ratio. R. of the minimum stress to the maximum stress (R= S,,,/ S,,,). If the stresses are fully reversed. the stress ratio R becomes -I: ii the stresses are partially re\,ersed. R becomes a negatke number less than I. If the stress is cycled between a maximum stressand no load. the stress ratio R becomes zero. lithestress isqcled bewssn two tensile stresses. the stress ratio R becomes a positive number less than I. .A stress ratio R oi I indicates no variation in stress. and the test \\ould becomea sustained-load creep test rather than a iatigue test. For carbon and lo\{-allo! steels. S-.Vcur\es typically halea fairI> straight slanting portion at low cycles changing into a straight. horizontal line at higher cycles. with a sharp transition between the two. An S-.I’cur!e usually represents the median life for a given stress-the life that half the specimens attain. Scatter of fatigue li\es can cover a \ery ibide range.
Source Metals Handbook.%h Park OH. 1978. p 667
Edann.
Volume I. Proprrr~rrnnd
Srlei~~on. lronsand Sreelr. .Amcrwan So&t!
ior hlctals. hlcr&
28
1-2. S-N Curves Typical for Medium-Strength Steels
100 Fracture region (all specimens fractured)
80
s:
rn
Fatigue - Iraclure band
lii rz: CIl ~
70
ViS:
~
i!
'"
CIl
0;-
:: (; 0_ .,,g
g>:;
c~
~-'
Q;
60
Finile-Iife region (no specimens fractured)
50
II
..
40 30
0..
Infinile-lile region
Fatigue limit
20 10 O~
10° (1)
_ _-L-_ _-L-_ _- L_ _- ' - - L - _ - L_ _- L_ _- L 10' (10)
103
(1,000)
10' (10.000)
_
10' 10 6 107 (100,000) (1.000.000) (10,000.000)
Number of cycles to Iracture
S·N curves that typify fatigue test results for testing of medium-strength steels.
As an explanation, if the single-load fracture strength of the specimens is considered to be 100 percent, for purposes of illustration this is the starting place, for the specimens can sustain no higher load without fracture. If ten specimens are fractured, the results are placed as points at the top of the left axis at one load application. Intuitively it is known that if the maximum load (or stress) is lowered to 90 percent of the tensile strength, it will require more than one load application to fracture the specimens. The ten points shown in the diagram at 90 percent represent the possible life to fracture of each of the ten specimens. Because the scale is logarithmic, the points appear to be relatively close, but in fact the scatter in life from longest to shortest is on the order of more than 2 to I. At this high stress, plastic deformation of the test specimen is likely to be great, such as in bending a paper clip or wire coat hanger to make it fracture. Actual parts are not intentionally designed to operate in this regime, 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, they will run longer with a fatigue life scatter of perhaps 3 to I, which is not unusual, even for theoretically identical specimens (which, of course, they are not). When the load is dropped to 70 percent, the lives get longer and the scatter in fatigue life increases to perhaps·5 to I. Again dropping the load, now to 60 percent of the single-load fracture strength, the fatigue lives again increase, as does the scatter from longest to shortest life. Invariably, in actual fatigue testing, there is at least one specimen that inexplicably fractures far earlier than any of the others in the same group. One such specimen is shown at the 60 percent level fracturing at about 150 cycles, while the other supposedly "identical" specimens or parts had lives of from about 1,000 to 10,000 cycles. The cause of such an early "anomaly" is often sought in vain, although it is possible that some metallurgical reason, such as a large inclusion on the surface, might be found. Frequently, this lone early fracture specimen is simply ignored. Dropping now to 50 percent of the single-load fracture strength, the fatigue lives increase dramatically, as the S-N curve starts to flatten out. This flattening out is characteristic of ferrous metals oflow and moderate hardness; many nonferrous metals and some very high-hardness ferrous metals tend to continue their downward path at very large numbers of cycles. Now, the problem is when to stop the tests. The test
1-2. S-N Curves Typical for Medium-Strength Steels (continued) machine will be needed for another test specimen after a very long test time, depending upon the rate ofloading, or cycles per minute. If ten million is selected as the end point, the test must be stopped at that figure even if a specimen is unbroken, and the point shown with an arrow pointing to higher values, for it did not actually fracture. Frequently, five million, or even one million, cycles is selected as the end point, depending upon the metal, purpose, and urgency of the tests. For example, five hundred million cycles is sometimes used in the aluminum industry. The region below the lowest portion of the S-N curve is called the infinite-life region, because specimens that are tested at stresses below the curve should run indefinitely; that is, they should have infinite life. The leveling of the S-N curve is the fatigue limit, characteristic of ferrous metals but not of most nonferrous metals. However, the region to the left of the sloping part ofthe S-N curve is called the finite-life region, for at the higher stress levels the test specimens or parts will eventually fracture in fatigue. 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. Also, growing fatigue cracks must not be permitted to exceed the critical flaw size characteristic of the metal and the stress state.
Source: Donald J. Wulpi, Understanding How Components Fail, American Society for Metals, Metals Park OH, 1985,P 135-137
29
30
1-3. S-N Diagrams Comparing Endurance Limit for Seven Alloys 100,000
...........
90,000
\
1.1 ao'tLerMcarOon.sttel, Oil quenchedan 1 drawn'
~~
~
80,000
K"'!>
~,
~~e>/
10,000
l'
'~6QOOO
n fir J II, I,
~.
r;
....
£
r-,
c-J.-
40,000
RC'-1,
'x
e:
" 30POO ~
~ '! ts« as rolled
:l,fio..L....!. I C
20,000 0.. ~...Q'
I
.1
104
rr
•
III
.I II. J.I III . I I,. I II.
Not:
.
<:0,"",;." ~ I· v.".,·c ~
i?o;C:P~ l'fo~"ea-lea
~aJfn
r- ~~n
10,000
a
!
dJ.S.Jpe,.l. cen/-carbon 'steel O'i''7ve", ("o$e
0.
.t50.000 ~
'wI. 0,7quenchedolnd -
I
I
(nd(cO/fis flf!f,mfn ~/c.;"lot rUf(furr
10 5 106 10' Number of c~c1e5 for rupture.Ioq scale
I
108
Typical S-N diagrams for determining endurance limit of metals under reversed flexural stress.
To determine the endurance limit of a metal, it is necessary to prepare a number of similar specimens that are metals tested, and for most nonferrous metals, the S-N diagrams become horizontal, as nearly as can be determined, for values of N ranging from 1,000,000 to 50,000,000 cycles, thus indicating a well-defined endurance limit. The S- N diagrams for duralumin and monel metal do not indicate well-defined endurance limits. The first specimen is tested at a relatively high stress so that failure will occur at a small number of applications of stress. Succeeding specimens are then tested, each one at a lower stress. The number of repetitions required to produce failure increases as the stress decreases. Specimens stressed below the endurance limit will not rupture. 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. Such diagrams are called S-N diagrams (S for stress, N for number of cycles). In general, the S- N diagrams are drawn using semilogarithmic plotting as shown in the above diagram, which presents the results for various typical materials.
Source: Fatigue and Creep Tests of Metals, P 220
1-4. Steel: Effect of Microstructure I. 0 r--
- - - - - - - - - - - - - - - - ----,
" 0.9 "i :--:.. 0.8 '<:: "i 0.7 o
'e 0.6
ec 0.5 e 0.4 :>
-g w
0.3 0.2 0.1
o Effect of steel microstructure on endurance ratio.
One of the more extensive investigations on influence of microstructure was conducted by Cazaud. The results of some of his work are summarized in the above bar chart. His data confirm that 0.5 is a conservative number; he found ratios varying from 0.55 to 0.62 for highly tempered martensites. These data were also for steels in the 0.40% carbon range. When untempered martensite is included, the total ratio range is from 0.26 to 0.62. Untempered 0.40% carbon martensite is about 55 HRC. Above 40 HRC, factors other than microstructure become more significant, especially nonmetallic content and residual stress. Many believe that tempered martensite gives optimum fatigue properties. However, much of the early work was with medium-carbon steels with intermediate hardnesses. Only limited data are available for other structures, including low-carbon martensites. Borik and Chapman determined the endurance limit of bainite and martensite in the range 36 to 61 HRC. They used 5ll00, a 1.00% carbon steel. They concluded that above 40 HRC, bainite had better fatigue properties at the same hardness than did martensite, whereas below 40 HRC the reverse was true. They explained the results in terms of carbide morphology and distribution. Below 40 HRC, the carbides in the martensite are spheroidal. Above 40 HRC, the carbide associated with the bainite was very fine and well-distributed, but below 40 HRC the carbides had a "pearlitic mode," which was less favorable in resisting fatigue.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure. American Society for Metals, Metals Park OH, 1979,P 77
31
32
l-5.
Steel: Influence
of Derating
Factors on Fatigue
Characteristics
Derating factors for influence of surface condition on fatigue. The graph above gibes C, factors for various surface conditions. It should be obvious that these factors are approximate. since it is impossible to represent such variable conditions by a single cur\e. C,. the size factor. is significant. Earl) work b> Horger firmly established that large-diameter samples of the same metallurg) were not asgood in bending fati@e as weresmall samples. In the presence ofa stress gradient. as in bending. a larger volume of metal is subject to high stress in a large part than in a small-diameter part. Since a large volume is subject to maximum stress. there is a higher probability of a critical-size nonmetallic inclusion to be in that volume. The fatigue properties established by testing large specimens are thought to represent the lower bound for a large number of small samples. Sinceaxial tests. b> their nature. test fairly large volumes at maximum stress. they also gibe lower-bound results. C,, is usuall) taken at I.0 for diameters less than O.J’inches and 0.9 for diameters between 0.1 and 3.0 inches. 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. The 0.5 relation for S,and S,, is onl! reasonably accurate in the low and intermediate hardness ranges because of limitations related to microstructure. nonmetallic-inclusion content. and carbon content 31 higher hardnesses.
Source: 0. H. Brcen and E. hf. b’ene. “Fatigue in hlashines and S~ruc~urcs~Ground American So&t) for Metals. hlerals Park OH. 1979. p 72
Vshlslcs. ‘. in Far~gue and hlcroaucture.
l-6.
Steel: Correction
Factors for Various
Surface
Conditions
Value for loading
in
Factor
Bending
K, .
1.0
0.58
0.9(a)
1.0
1.0
1.0
Kd, where: d c 0.4 in. 0.4 in. < d 5 2 in. . . . K. . . . . . . . . .
Torsion
33
Tension
0.9
0.9 1.0 From chart above (aI A lower value 10.06 to 0.85) may be used to account for known or suspected undetermined bending because of load eccentricity.
Correction factors for surface roughness (kJ, type of loading (14). and poti diameter (KJ, for fatigue life of steel ports.
Comparative effects of various surface various levels of tensile strength.
Source. Metals Handbook.9rh Parh OH. 19% p 6’1
Edlrlon.
Volume
I. Properr~erand
conditions
Selcc[ia,n.
on fatigue
Ironsand
Srscls.
limit
of steels at
American
SO~ISI)
ior hlclalr. hleralr
34
l-7.
Fatigue
Behavior:
Ferrous vs Nonferrous
Metals
S (stress)--N (cycles to failure) curves. 4. ferrous metals; B. nonferrous metals. S, is the endurance limit.
Traditionally. the behavior of a material under conditions of fatigue has been studied by obtaining the S-.‘l’cur\es (see above), where S is the stress and .E- is the number of cycles to failure. For steels. in general. one obsenes a fatigue limit or endurance limit (curve A above) which represents a stress level below which the material does not fail and can be cycled infinitely. Such an endurance limit does not exist for nonferrous metals (curve B above). The relation between Sand ,V. it must be pointed out. is not a single-ralue function but serves to indicate a statistical tendency. Up until the 196Os, almost all fatigue failures. and consequently all the research in the field, was confined to moving mechanical components (e.g.. axles. gears, etc.). Starting in the late 1950s. entire structures or very large structural elements (e.g.. pressure Lessels. rockets, airplane fuselages, etc.) have been studied and tested for fatigue. This can beattributed to the use of materials such as high-strength alloys. together with the advances in the fabrication technolom. resulting in monolithic structures meant to undergo high cyclic stresses in service. It is this class of materials which has shown catastrophic failures in fatigue. and it is for this kind of material that fracture mechanics is being applied. with considerable success. to fatigue problems.
Source: Marc Andrt Meyersand Krishan Engleuood Chfk NJ. 198-l. p 689
KumarChauls.
hlechanlcal
hle~allurg):
Prmaplesand
.Appllcaionr.
Prcm~ce-Hall.
Inc..
l-8.
Comparison
of Fatigue Characteristics: Aluminum Alloy
Typical bending (R = -1) fatigue rous and nonferrous metals.
curves
Mild Steel vs
for fer-
Here it is noted the lack of the “knee” for the aluminum allo) compared with steel: that is. the point on the cun’e where the number of cycles to failure becomes a straight line-essential11 infinity.
35
36
1-9. Carbon Steel: Effect of Lead as an Additive 120~-...,...--r----r--r---,---,---.-----,---.--...,.-----,
IOo/-----1I--+----1--+-----+~~ V>
n.
o BOI-------t---t----r-----: o o
Fatigue limit of leaded and nonleaded alloy steels as a function of ultimate tensile strength.
Lead is often added to steels to improve machinability, although usually at the cost of a minor (usually) loss in mechanical properties,. The interrelationship of lead additions with tensile strength and fatigue limit is summarized in the above graph.
Source: George M. Sinclair, "Some Metallurgical Aspects of Fatigue."in Fatigue-An Interdisciplinary Approach. John J. Burke, Norman L. Reed and Volker Weiss, Eds., Syracuse University Press, Syracuse NY, 1964. p 68
1-10. Corrosion Fatigue: General Effect on Behavior
t
Fatigue without corrosion Fatigue limit
Logarithm of number of cycles needed for fracture _
Effect of alternating stresses with and without corrosion.
If a specimen is subjected to alternating stress (tension and compression in turn) over a range insufficient to cause immediate fracture, gliding may occur within some of the grains, but when the disloca tions reach a grain-boundary they are halted, retracing their movement along the gliding-plane when the stress is reversed. If the material were ideal, it might be hoped that the dislocations would merely move to and fro along the plane, and that no damage would result. In practice a large number of cycles can be withstood without apparent damage, but in material as we know it, slight irregularities will prevent smooth gliding indefinitely, and roughening along the original glidingplane will make movement difficult, so that gliding will then start on another parallel plane. In the end, bands of material will have become disorganized, and ultimately one of two things must happen: (1) if the stress range is low, gliding will cease altogether, the only changes still produced by the alternating stress being elastic, (2) if it exceeds a certain level (the fatigue limit) the gliding will become so irregular, as to cause separation between the moving surfaces, first locally, producing gaps, which later will join up into cracks. Thus above the fatigue limit (after a time which is shorter at high stress ranges), there will be failure;
below the fatigue limit, the life, in absence of corrosion, should be indefinitely long as shown above. In the presence of a corrosive environment the situation will be different. Disorganized atoms along a gliding-plane may require less activation energy to pass into a liquid than more perfectly arrayed atoms elsewhere; certainly, while the atoms are in motion along a gliding-plane, preferential attack may reasonably be expected even below the fatigue limit. This means that there is no "safe stress range" within which the life should be infinite. It is, however, convenient to determine an endurance limit-namely, the stress range below which the material will endure some specified number of cycles (the number must be stated). It should be noted that, although stress-corrosion cracking is often intergranular, corrosion-fatigue cracks are usually transgranular, following glidingplanes inclined at such an angle as to provide high resolved shear stress. There are exceptions to both rules. Whitwham, studying corrosion-fatigue cracks on steel, found that, although mainly transgranular, they followed grain-boundaries for short distances, where such boundaries chanced to run in a convenient direction.
Source: Ulick R. Evans, An Introduction to Metallic Corrosion, 3d Edition, Edward Arnold (Publishers) Ltd and American Society for Metals, Metals Park OH, 1982, P 160
37
38
1-11. Effect of Corrosion on Fatigue Characteristics of Several Steels l00r------------,
'Copper' steel
l' 09')(, C Steel
80
:c u
O 104 c. 100 ,e Ii;
-
40
~
'0 :> 0
:E "'"e e
~
'E
"
(/)
:c u
105
106
107
.5
g-
lOB
ill
80
ill
40
20
104
105
106
107
lOB
:E "'"<: e ~
"R"."~
~
:> 0
60
l!!
"
104
105
106
.1£
E
Chrome-
"
vanadium 20 steel
(/)
O'BB')(, Cr,0·14% Vo,0'46')(, C lOB
.,
0 104
105
106
25 104
105
106
107
\4
50
,,~
chromium
i:
107
lOB
30 104
105
106
107
N = Cyclesto fracture 1I0g scale)
---Denotes testsmadein air 6------
"
"stream of fresh water II
lOB
Tensile: 81000 Ib.lsQ,ln.
Tensile: 150600 Ib.lsq,in.
107
35
steel 27')(, Cr.0'2')(, C
(hardened and temperedl
(hardened and tempered I Tensile: 65700 lb.Zsq.in,
0 103
~
40
'E (/)
:E "'e"
. ~ 66
c.
'0 i:1<:
:> 0
0 103
12·9')(, Cr,0·11')(, C Tensile: 89600 Ib./sq.in
45
Ii;
,e
0
i:1<:
20
w "0
e ill
(hardened and tempered)
0·14')(, C,O'9B')(, Cu Tensile:61500 Ib./sQ.in.
g-
ii.5
,e
20
.5
60
g"c.
~
40
(annealed) Tensile: 103500 Ib.lsq.in.
saline riverwater
Typical curves showing the number of cycles needed to produce fracture at different stress ranges in absence and presence of corrosion.
Two main procedures are available for corrosion-fatigue tests: One-stage tests. Here the corrosion fatigue is continued until breakage. The logarithm of the number of cycles needed to produce breakage is generally plotted against the stress range, as in the above curves selected by Gough from McAdam's experimental data. Two-stage tests. Here the corrosion fatigue is interrupted after a definite number of cycles, and the residual strength is estimated by measuring either (a) the endurance limit in the absence of corrosive influences (i.e., the stress which can be withstood for some definite number of cycles, (b) the number of cycles needed to produce fracture in the absence of corrosive influences at some definite stress, (c) the tensile strength, or (d) the shock resistance (Izod number).
Source: Ulick R. Evans, An Introduction to Metallic Corrosion, 3d Edition. Edward Arnold (Publishers) Ltd and American Society for Metals, Metals Park OH, 1982, P 165
lOB
1-12. Steel: Effect of Hydrogen on Fatigue Crack Propagation
I
PR PAGAIION LIfE
I -,---_. I i
1+--.,--,---,--..c--,...-,-,--.-.+---.--.---.--.,..-,-,....,.......,r+--...,.----,.---.. 1
6 7 0 9 0
S-N type of fatigue curve.
In the majority of all cases, the external load changing with time, whereby low frequencies «10- 2 Hz) have the highest practical importance. Under these circumstances a structural component can be subject to fatigue which is conventionally described by an S-N curve relating the cycle life, N, to applied stress, S, as in the above chart. 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. The number of the cycles corresponding to the endurance limit presents initiation life primarily, whereas the number of cycles for crack initiation at a high value of applied stress is negligible. Consequently S-Ntype data do not necessarily provide information regarding safe-life predictions in structural components. Particularly, if the structure contains surface irregularities different from those of the test specimens, these are likely to reduce or even eliminate the crack initiation portion of the fatigue life.
Source: M. Kesten and K.-F. Windgassen, "Design of Equipment to Resist Hydrogen Fatigue Service," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Pressouyre, Eds., American Society for Metals, Metals Park OH, 1982, P 390
39
40
l-1 3. Relationship
of Stress Amplitude
(a) Finite life ASR diagram. showing S-.Y diagram, showing life prediction stress.
and Cycles to Failure
R = -I equivalent stress for R = 0.6 loading. (b) for R = 0.6 loading using R = -I equivalent
The ASR diagrams normally use theendurance-limit fatiguestrength value. but substitution of the fatigue strength at specific finite li\es can also be used (see chart a abole). Life estimations from the diagram can be done using such information as is shottn in charts a and b abo\,e. Here the hno\\n stress range at some Rvalue is conierted to an equivalent completely reversed (R= - I) stress. and thisequkalent stress is applied to the matefor the life estimate. rial’s R = -I S-.Vcur\e Designers have the ability to calculate the component’s stresses using classical formulas or the computer-based finiteelement-analysis (FEA) techniques. Both of these methods examine the elements for the maximum stresses that are normall> in the areas of a discontinuity. or stress concentration.
Source. D. H Brscn and E hl Hens. “Faugue m hlachlner and StrucrtmpGround Metals Park OH. IYTY. p 6’ Amencan So&l! ior hkrals.
\ ehlcles.” in Fatigue and hlwrosrruc~ure.
1-14. Strain- Life and Stress- Life Curves
2Nf
Reversals to failure (lag scale)
Strain-life and stress-life curves,
Fatigue damage is caused by cyclic plastic strain, and consequently, the fatigue life should be related to the plastic-strain amplitude. Coffin and Manson independently proposed a relationship between the plastic-strain amplitude and the cycles to failure of the form: AEp/ 2 =
EJ (2NJ),
where Ej is the fatigue-ductility coefficient, 2NJ is the number of reversals to failure, and c is the fatigue-ductility exponent. Their equation is very similar to the Basquin equation relating the elastic-strain or true-stress amplitude to the number of load reversals to failure: I::.E,E/2
= aa = aJ (2NJ)b
where I::.E,/2 is the elastic-strain amplitude, E is the modulus of elasticity, his the fatigue-strength exponent, and is the fatigue-strength coefficient. A schematic representation of these relationships and their superposition is shown in the above diagram. The summation curve is analogous to the stress-life, Wohler diagram, if the strain amplitudes are replaced by their respective stress amplitudes. The intersection of the Basquin and Coffin-Manson plots is normally defined as the transition between high- and low-cycle fatigue. Consequently, the regime of low-cycle fatigue depends on the properties (for example, the ductility) of a particular material.
a;
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, p 211
41
42
1-15. Fatigue Plot for Steel: Ultrasonic Attenuation vs Number of Cycles
0.264 0.249
E
~
CD
0.232
:5! t:I
z 0.216 0 ;::: 0.200 «
ADDITIONAL PULSE OBSERVED
:::> Z
~
«
0.184 0.168
0.4 dBATIENUATION CHANGE OBSERVED
l
l
0.152 5 10.8xl0
Typical plot of ultrasonic attenuation versus number of fatigue cycles for steel.
Joshi and Green determined the attenuation coefficient IX for longitudinal bulk waves in aluminum and steel at 10 and 5 MHz, respectively. The measurements have been performed in a pulse-echo mode, with the acoustic pulse reflected at the back surface of the material. The above chart shows their results obtained on cold rolled steel bars. The attenuation started to increase at about 6 X 105 fatigue cycles (65% of fatigue life). At roughly 7.5 X 105 cycles (85%), an additional pulse was observed, arriving earlier than the one reflected from the back surface. Results are interpreted in terms of a series of microcracks being formed, probably at the surface. As soon as the microcracks are sufficiently deep, they will change the bulk attenuation. As soon as a macrocrack has been formed (by coalescence of microcracks), it will reflect part ofthe pulse. After that, the attenuation is primarily determined by the transmission coefficient of this single crack. Thus, the attenuation curve (versus fatigue cycles) becomes discontinuous, as may be noticed in the above chart.
Source: O. Buck and G. A. Alers, "New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals. Metals Park OB, 1979. p 135
2-1. Typical S-N Curve for Low-Carbon Steel Under Axial Tension 28
I
26 ~
c
~
TYPICAL FATIGUE CURVE FOR M.S. UNDER REPEATED AXIAL "TENSION (f MIN. = 0)
1.
~ 24 I
rc)(
I
«
~ 22
I
't-
'-'
III III uJ 0::
J-
i 20
\
III
0 W .J 0. 0.
18
16
14
t
\
I~
--
~ --fATIG-U-E-LlMi.:r-(REPEATEO- TEtlSION)
o
I
I
I
2
4
G
~
8
10
\2.
NUMBER OF STRESS APPLlC",.\ONS- MILLIONS
The term "fatigue" refers to the failure of metals from repetitions of stress rather than from a single application, as occurs for example in a simple tensile test or with a brittle failure. The value of the stress necessary to cause failure of a material from fatigue is lower than its nominal tensile strength. For example, a sample of mild steel may have a maximum stress of 27 t.p.s.i. when subjected to a single application ofload as in an ordinary tensile test. If, however, a stress of say 25 t.p.s.i. is applied repeatedly to the same material, failure will not take place until this has been done a certain number of times, while at a lower stress still, the number ofload cycles required to cause failure will be even greater. If testing is continued in this manner, a stress value will ultimately be found at which fracture will not occur, no matter how many stress repetitions are applied. This value is known as the fatigue limit of the material. If the results from such a series of tests are plotted, a graph such as the one above will be obtained, the curve tending to run parallel to the abscissa after approximately IOmillion cycles (for steel), the corresponding value ofthe stress being known as the fatigue limit. Under conditions of repeated tension the value of the fatigue limit for the above mild steel which has a tensile strength of approximately 27 t.p.s.i. would be of the order of 16 t.p.s.i. If the same steel was tested under conditions of reversed bending stresses a value of the order of± 12 t.p.s.i. may be found. It must also be pointed out that where corrosive conditions operate in addition to fluctuating stresses, failure from "corrosion-fatigue"may occur and, in these circumstances, the concept of a fatigue limit does not apply, since if the stress applications are continued for a sufficient number of times, ultimate failure will occur. Further, most nonferrous metals and alloys do not possess a fatigue limit.
Source: F. R. Hutchings, "Fatigue Failure of Components of Lifting Machinery," in Failure Analysis: The British Engine Technical Reports, F. R. Hutchings and Paul Unterweiser, Eds., American Society for Metals, Metals Park OH, 1981, P 344
43
44
2-2. AISI 1006: Effects of Biaxial Stretching and Cold Rolling
..
8
C'l
Eoff
cP
b
.. 0.2
)(
~
..r
s
00.4 a 0.6
6
"j
c
:::> f-
:::i
Do
::l!:
oct z
4
« a: lii ..J
oct
2
f-
0
a
f-
10 3
104
loS
10 6
REVERSALSTO FAILURE, 2N,
..
C'l
b
8
)(
Eeff
~
.. 0.2
s
00.4
..r
w·
c
a 0.6
6
:::> f-
:::i
Do
::l!:
oct z
« a:
4
f-
Vl ..J
oct
f-
0
f-
2
Runouts
r 103
104
loS
10 6
REVERSALSTO FAILURE, 2N,
Strain-life plots for two modes of deformation for 1006 steel.
Plots in the top chart are for biaxial stretching; those in the bottom chart are for cold rolling. Included is the data band for the undeformed material. The effect of balanced biaxial stretching on fatigue life was as follows: at large strain amplitudes (/:;.EI/2~ ~ 2.5 X 10-3 ) , the fatigue life remained approximately the same or decreased slightly when compared to that ofthe undeformed material; in contrast, at small strain amplitudes the fatigue life increased as a result of the prior deformation. After cold rolling, the fatigue life was approximately the same as in the undeformed material at large strain amplitudes (short lives) but it was longer at small strain amplitudes (long lives). Thus, unlike BBS, CR appeared to cause no reduction in fatigue life at short lives. Another difference between the two deformation modes was that the scatter ofthe data was larger after BBS than after CR. Thus, BBS was somewhat more detrimental to the fatigue life than CR.
Source: John M. Holt and Philippe L. Charpentier. "Effect of Cold Formingon the Strain-Controlled Fatigue Properties ofHSLA Steel Sheets," in H'Sl.A'Stccls-c-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 217
2-3. AISI 1006: Weldment; FCAW, TIG Dressed
SAE-1006 R' 0.1. t' 0.13" 13.3mml --~--
Smooth Specimen --<>-- TIG-Dressed
_.--fr-.- As-Welded
_
00 0 - - - - 0 0 roo 6_ --rrtr-lS" i:J_ _ tr 6 ~D;!.~_ 0 0 0 -o 0 .-._ oro ~ '-.0 0-. o DO ~._._
._....g 0
'-''''lJ..._ '--0
10 5
..............
10 6
NT' CYCLES TO FAILURE Fatigue strengths of FCA W/TlG- dressed joints compared to those without TlG dressing for AISI I006steel (unwelded). The improvement in fatigue provided by TlG dressing tbe welds is obvious.
Source: Kon-Mei Ewing, Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens. "Weld Fatigue ofTlG-Dressed SAE-98QX HSLA Steel,"in HSLA Steels-Technology & Applications. American Society for Metals, Metals Park OH. 1984,p 557
45
46
2-4. AISI 1006: Weldment; Shear Joints
~. or
CJ)
•
•
•
.
(a,.-32KSI \; ur= 32 KSI
SAE 1006 LAP-SHEAR WELDS Kfmax = 2.77 I R = 0.1 • EXPERIMENT PREDICTION
10 5
NT
106 I
CYCLES
Total fatigue life predictions and experimental results for FCA W, AISII0061ap-shear joints. Note that the results and predictions compare closely.
Source: Kon-Mei Ewing. Pei-Chung Wang, Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue ofTIG-Dressed SAE-980X HSLA Steel," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984,p 562
47
2-5. AISI 1006: Weldment; Lap-Shear Joints
~
(f)
10 1
0
•
•
•
.
€~"-'2KSI
;;;
CTr =32
SAE 1006 LAP-SHEAR WELDS
a. ~
10
2
KSI
Kfmax = 2.77. R = 0.1 EXPERIMENT PREDICTION
•
101
NT,
CYCLES
Total fatigue life predictions and experimental results for FCAW, AISI 1006 lap-shear joints. Here, the prediction and actual results are very close.
Source: Kon-Mei Ewing. Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue ofTIG-Dressed SAE-980X HSLA Steel, "in HSLA Steels-Technology &Applicalions, American Society for Metals, Metals Park OH, 1984,p 562
vi
48
2-6. AISI 1015: Effect of Cold Working
0.20
5.0 . E ~ E
.x:
4.0
0.15
<, M
E E
0-
'"
M
I
0
3.0
I-
:E
ui
-O.I0:J
I
a: a: 2.0
uJ
::J
'"
lLJ
::
0.05
I
u,
1.0
oL--!------~---_:::_---_::::__'O
o
20
40
DEGREE OF COLD-WORKING,
60
%
Comparison of effects of cold working on wear rate and fatigue limit of fully annealed 0.15%C mild steel. Wear was determined in sliding between the end surfaces of cylinders at a speed of 0.56 mjs under the loads .:82 N, ():124 N and 0:147 N in machine oil. Fatigue limit (.) was determined by reversed bending fatigue tests of notched plate specimens 25 mm wide and 4 mm thick having a central hole 1.5 mm in diameter.
Attempts have been made to determine effects of cold-working on the resistance to wear and fatigue of a O.15%C mild steel. Fully annealed material was then cold-worked to different degrees and the specimens were machined from it. Wear experiments were conducted in a rotating cylinder machine as described above with a machine oil as the lubricant. Care had been taken to avoid the effects of work hardening during machining by electrolytically polishing the sliding surface. Reversed bending fatigue tests were carried out by using notched test pieces of the same material. The wear rate and the fatigue limit are compared with the degree of cold-working in the above chart, which shows a definite correlation.
Source: Yoshitsugu Kimura, "The Role of Fatigue in Sliding Wear," in Fundamentals of Friction and Wear of Materials, David A. Rigney, Ed., American Society for Metals, Metals Park OH. 1981, P 215
2-7. A533 Steel Plate: Fatigue Crack Growth Rate Stross-intensity factor rango, .0. K, ksi • in. 1/ 2
50
10- 2 10- 4
a <9':!
10- 3
~
~ E E
Z
~ co
...
01 00 I 1
Region 1: slow crack growth
I Region 3: I rapid I unstable I crack growth I I I I I I
10- 4
",'
1!
ie at
~
eu
...e
10- 6
Region 2: power-law behavior
'"
U
~ .5 Z
~ co
...
ti
10- 6
l!
~ e ...u'" l!u
...l!!
I
~
'" :iE
10- 6
10-7
~
'" :iE
10- 6
Stross-intensity factor rango, .0. K, MPa . m 1/2
Fatigue crack growth behavior of AS33 steel. The material was ASTM AS33 B-1 steel, with a yield strength of 470 MPa (70 ksl), Test conditions: R= 0.10; ambient room air; 24°C (75 OF).
The general nature of fatigue crack growth and its description using fracture mechanics can be briefly summarized by the example data shown in the above chart. This figure, based on the work of Paris et al, shows a logarithmic plot of the crack growth per cycle, daj dN, versus the stress-intensity factor range, t::.K, corresponding to the load cycle applied to a sample. The da] dN versus t::.K plot shown is from five specimens of ASTM A533 B-1 steel tested at 24°C (75 OF). A plot of similar shape is expected with most structural alloys; the absolute values of daj d N and t::.K are dependent on the material. Results of fatigue crack growth rate tests for nearly all metallic structural materials have shown that the da I dN versus t::.K curves have the following characteristics: (a) a region at low values of dald N and t::.K in which fatigue cracks grow extremely slowly or not at all below a lower limit of t::.K called the threshold of t::.K, t::.K,,,; (b) an intermediate region of power-law behavior described by the Paris equation:
~=C(t::.KJ' dN
Source: J. H. Underwood and W. W. Gerberich, "Concepts of Fracture Mechanics." in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH. 1982. P 18
49
50
2-8. A514F Steel Plate: Fatigue Crack Growth Rates
20
MPavm
100
100
20
10'
A514F CON Quality
'(J '
I /; /,.
., 100
10
z!.
I
~ 10"
<
A514F CaT Quality
~
~ E E
lS----1
u-
-::'O··
15-----·< TL-- ---- .. - --- -_ :
~I=-:-:=:':::- i l'.K. ksl\ ii1.
'00
Plots of fatigue crack growth rate versus range of stress intensity factor (best fit lines) for A514F plates.
The increased isotropy in the CaT over the CON steels is evident with the through thickness (ST, SL) orientation having the fastest growth rate in the CON steel and showing the greatest improvement by CaT.
Source: Alexander D. Wilson. 'The Effect of Inclusions on the Properties of Constructional Steels." in Wear and Fracture Prevention. American Society for Metals. Metals Park OH, 1981. p 196
2-9. A514F and A633C: Variation in Fatigue Crack Growth Rate With Orientation
5.10-'
~Kof 50 kst/iil (55MPa.frii)
CON~
A633e
CaT
A633C
0
CON~
A514F
caTD 1110"' 3.10-'
~§~~
ICaT I 4.10·'
A514F 8x10"
12110"
16.10"
20110-'
daldN, inches/cycle Comparison of fatigue crack growth rate variation with orientation for A633C and AS14F plates at two tJJ( levels.
These data show that the CaT improvement in FCP growth rate takes place only at higher L1Klevels. Additionally, this figure indicates that there is a more substantial enhancement in FCP behavior for ASI4F. Also there generally appears to be more anisotropy in the ASl4F steels of both quality levels. It has previously been shown that higher strength level steels tend to be more adversely affected by inclusions associated in groups, such as present in CON steels.
Source: Alexander D. Wilson, "The Effect ofInclusions on the Properties of Constructional Steels, "in Wear and Fracture Prevention, American Society for Metals, Metals Park OH, 1981, P 197
51
52
2-10. A514F: Scatterbands of Fatigue Crack Growth Rate MPavm
20
:------ .....,
10. 0
I I
100
I
, ,: , ,,, ,
I
I
I
,
I
I
10"
I
I
I
I
I
I
I
I
I I I I
I
I ~
~
I
}
A514F
E E
2'/.ln(57mm)Gage CjCON ClCaT
10"
10·' L-_ _-----''--_-'------'_'-'-'--'---'-L-_ _-----' 10 100
6 K , ksi\/fil.
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.
Source: Alexander D. Wilson, "The Effect oflnclusions on the Properties of Constructiona ISteels,"in Wear and Fracture Prevention, American Society for Metals, Metals Park OH, 1981,p 197
2-11. A633C Steel Plate: Scatterbands of Fatigue Crack Growth Rates 20
MPavm
100
10-'
10-' GI
I
GI
'fi 10"'
,5
4In(102mm)Gage
z
C=JCON
~
E
DCaT
10-'
I
A633C
10-'
L-_ _---'-_---''---'------'-----'----'--'--'---'--_ _------"-'
10
,c.. 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.
In this presentation the generally faster FCP growth rates for the CON steels at higher 6.K levels are displayed, as well as the improved isotropy of the CaT steels.
Source: Alexander D. Wilson,"The Effect of Inclusions of the Properties of Constructional Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OR, 1981, P 196
53
54
2-12. Low-Carbon Steel Weldment: Effects of Various Weld Defects 100
I
80
-
60
r-
... 1"-1-
40
0
-,~
- --
---
0
...
uv
o VV -in .0/.
-
'."C -
8 f-
6 fI
10'
"
data bank
~
.,
~
'flo
9-
LC?cation of failure: o Plate or weld edge • Porosity in weld ... Slag near surface b. Slag at midthickness V Lack of penetration
10
I
r----... • r-9 ·0 Data bank ;--I-- r-- mean f..' curve
IA
Low-carbon steel reinforcement off
~ ch
I,t:
I
. --
~-
_C
ri if.. ...ll ...
20
on
I
+ 1 standard deviat ion
....c
.,
OJ"
e»
~
I
I
poli:he~ Plai~ pl~te
V
V
111
11 V
11
I~
1<:;1-
I
2
4
6
B 10'
2
4
6
Fatigue life, cycles
8
10
•
2
4
6 8
10
,
S-N curves showing effect of various weld defects on fatigue life of a low-carbon steel weldment, presented as a comparison with fatigue life of the plate.
Source: Metals Handbook, 9th Edition, Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park OH, 1983, P 848
2-13. Low-Carbon Steel Weldment: Effect of Weld Reinforcement and.Lack of Inclusions
-e
Reinforcement intact Reinforcement removed
;j!. "0'
0;
;:
"0 C ::J
g
(;
-----
-S
'"
~
40 _---""
___-~......=c-I_---~I""'-_~=__iI_---___l
----
1;; CI>
.,
::J
.'"
'"
u,
201_----1_----1--
Ol.--
o
.l.-
..l-
...J......
...l-
0.2
0.4
0.6
0.8
-J
1.0
Slag inclusion length, in.
Fatigue strength of a weldment containing slag inclusions as a percentage of the mean fatigue strength of a sound low-carbon steel weld.
Source; Metals Handbook, 9th Edition, Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park OH, 1983, P 850
55
56
2-14. Low-Carbon Steel Weldment: Effect of Weld Reinforcement and Lack of Penetration 100
\\ 80
60
\\
r-....
-,~'" <,
"
1'-, ..... ........... .....
R . fl. I em orcement mtact
-
~I
I
<, ~
t--.
A
100
oob cycles
--.- r--r-.r------<.
.................. ~ r--
40
I
Reinforcement remived
.i
\/ -:
~
/
-.
-.-
-
2 000 000 cycles 20
o o
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Lack of penetration half depth, in.
Fatigue strength ofa weldment containing lack of penetration as a percentage ofthe mean fatigue strength ofa sound low-carbon steel weld.
Source: Metals Handbook, 9th Edition. Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park OH, 1983, P 849
2-15. Low-Carbon Steel Weldment: Computed Fatigue Strength; Weldment Contained Lack of Fusion 100
;fi.
80
,,' Qj
\ ~
1\
\'\
'\ III
"
<,
s
-g 60
~
'0 -S '" ~ 40
r-, <,
....
Reinforcement intact I I ' I - . - Reinforcement rejOved
<, <,
"""-
" - -::::--- -~
....... ..........""'" ~ ~-- ~ b..
r-._ r--
Ii:
1 100 000 cycles
J
•
Ql
".,'"5.
LL
--
~.- ::-- ....t.
}
2 000 000 cycles 20
o
o
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Lack of fusion half depth, in. Computed fatigue strength of a weldment containing lack offusion as a percentage ofthe mean fatigue strength oCa sound lowcarbon steel weld.
Source: Metals Handbook, 9th Edition, Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park QH, 1983, P 849
57
58
2-16. Low-Carbon Steel Weldment: Effect of Reinforcement and Undercutting 100
80
*'
'0'
~
'0
5 g
60
...o
fia>
~
40
I
~
~~ ~ \~
\ [\
<,
~
~ .......
I" '-,......r-
10-. ~
""'-
r-; :::::-
Cl>
:::l
I
t::-I-- i'---
'
r-- 1-.
a>
co
--. "k
0.Q2
0.04
0.06
-
""'--. ~ t:- t--
•
r
20
a a
;00 000 cycles
r::: t:-- r--.r-- .r:;'/ to-.
';;
u..
I
. - Reinforcement removed
r': t::'- r---
t:
. fl.
""""- ern orcement Intact
0.08
0.10
t--
00 000 cycles
0.12
0.14
0.16
Undercut depth, in.
Effect of depth of undercut in terms of percentage of fatigue strength of a sound low-carbon steel weld.
Source: Metals Handbook, 9th Edition. Volume 6. Welding, Brazing, and Soldering, American Society for Metals, Metals Park OH. 1983. P 848
2-17. Low-Carbon Steel: Transverse Butt Welds; Effect of Reinforcement •
•
'I
........
600
•
r ,
'1-0---
, - - - , - - r ,-,.-.-,--""--"~J.------,---r---"I---y---r-T"""T""
"-
~"""""O
500
• •
x
ro
300
E (/)
200
100
0 0
•
~<, O~ ~REINFORCEMENT OFF .~.,-.-3 ~ Q.... o __•
•
~_
1.5
••
~ ••
2.3 - .
0
*• .•
-
-~-
REINFORCEMENT
Q..
't:..
ON
3'-............ ~--. t.: ..~ •• j ----t--~ ......... ...
h =3.8mm
r h
Tr==r?J
~
<,
.
<,
.'6~-------~ __ C ~. ~
•
•
-.r:~
CYCLES Influence of weld reinforcement on fatigue strength (R=O) of transverse butt welds of quenched and tempered carbonsteels, From these data it is evident that removal ofthe reinforcement (weld dressing) improves fatigue strength and fatigue life.
Source: Drew V. Nelson, "Fatigue Considerations in Welded Structure," in Proceedings of the SAE Fatigue Conference P-109, Society of Automotive Engineers, Inc., Warrendale PA, 1982, P 206
59
60
2-18. A36/E60S-3 Steel Plate: Butt Welds IOOr:---,----,.--..-r-.-rn"----r--r--r-OT,,""TT--..--r-r-,T"T,,,
80
600
6Or---====:=--_
400
300
40
CT, =0
200 20
10
100~ 7O:e
8
5O
CT,' +35 ksl
~
vi
6
/!l361 E60S-3 Double-V Bull Welds
4
~ma.'2.5I,
30
R=0, I = 5/8 in.
OCT, = + 35 ksl
S
20
S
2 10 7
10 N1 ' Cycles
Fatigue crack initiation life predictions and experimental results for ¥a-in. (16-mm) A36/E60S-3 butt welds.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981,P 114
2-19. A514F/E1·10 Steel: Bead on Plate Weldment
~14F/EII(j Begd On plgte
Kfmal •
31~1 DJ • 0.01In.•
R -0. t -112 ln.
- - . "".120kl' ---0 "" ·-120kli
200
200
.
100
-:---_-
vi
•
---
00 0 - 0 - - - _
l.
10 ~
vi
so
"'----
30
20 20
10
Total fatigue life predictions and experimental results for A514F /EllO weldments with tensile and compressive residual stresses.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 113
61
62
2-20. A36 and A514 Steel Plates: Butt Welded
514F / E 110 Bull Wold 10
~
8 6
100 ~ 70 ::f.
K,...3.15, 01- D.OIIn. (0254mm)
.-90·, , -60·, I -1/2 In.. (I2.7mml
50
--.,.s,
- - - ' , -0
~
..
s~s K,..... r
'"
~
30 20
10
NT' Cy(lu
Predicted effect of stress relief and stress ratio on A514/EllO butt weld fatigue life.
1~'r--~--r-r""""~"'--~-~~""""~"'--~-'-"""'~~"'6oo
400
60
300
40
R'O
20
A36/E60S-3 Bull Weld K,,"_ ·2.35, Ot' 0.01 In (O.254mmJ 10
8 6
.-90·,' -60·, 1-I/Zin,,(I27mm)
--.,-s), - - - w,-O
~-.
s~s K r_ .
30 20
' 10
10<
10' NT' Cycle,
Predicted effect of stress relief and stress relief and stress ratio on A36/E60S-3 butt weld fatigue life.
The results for the high-strength, quenched-and-tempered steels (upper chart), indicate that such materials can sustain high residual stresses which do not relax. The total fatigue life of such materials is strongly influenced by both residual stress (a,) and stress ratio (R). Stress relief or mechanically induced compressive residuals should be highly effective. An intermediate case is mild steel as shown in the lower chart. Mild steels can have appreciable residual stresses; but, since the transition fatigue life (N,,) is often very long (= 500,000 cycles), there are large amounts of plasticity at the notch root even at long lives (106 cycles); this notch-root plasticity tends to relax rapidly the notch-root residual and mean stresses with the result that N[is little affected forlives less than 106 cycles. The observed dependence of N» on stress ratio does, however, result in a predicted variation oftotal fatigue life with stress ratio R.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 112
63
2-21. A36 Plate Steel: Butt Welded 100 80
600
60 Zero Mean Slress
40 200 20
...;; VI
10 8
--\---====::::::::d Mean Slress Effects With No Rela.atian
~~
6
4
2
e
100 70 2:
30 A36 Butt Weld (HAl) KImox =3 I R-O a r =+35 ksi
20
10
IIO~
10' N
I
• Cycles
Mean stress relaxation behavior influence on fatigue crack initiation life (A36 HAZ material, K f = 3, R = 0, a r = +35 ksi (242 MPa) ).
Materials such as high-strength steels exhibit very little notchroot plasticity; consequently, a os may be larger than a r- The results obtained using the model agree with the experimentally observed behavior. The above chart shows the qualitative behavior of N[ predictions.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, pIli
64
2-22. Low-Carbon Steel Tubes: Effect of Welding Technique .,.--'
1·0
-----
0·8
'--'.-, "" ~
" '" .
,
- - - - f----- _._ •..• _ -
1-. - ' ,,""' ~,~'-. " I\. ....
"
"
i'..'..
'-'0
O·B
"
"-
,
, .,~:, ','~ ' ~~ ,~ ~/ ..,~ ~'/~
~
~
-~/
-
.,~
~
.-:~~ /.--~-~
I~'" V'~
%-. ~/.
..,~
~
'~r::-, '/8
~.,
0·4
f--------.--
~c
b-.,
-----_.-
A_
~.ij'/W
,-
~.
'~
~D
0·2
o
1
I
I
I I
I
I
I I
I
I
I I
I
I
I
I
I I
I
I
I
I
10 Fatigue strength of welded tubes: A - unwelded or welded without filler metal; B - helical welding (700 angle); C - longitudinal or helical welding (550 or 600 angle); D - helical welding (500 angle).
Source: R. V. Salkin, "Low Cycle Fatigue of Welded Structural Steels: A Material Manufacture and Design Approach," in Proceedings of the Conference of Fatigue of Welded Structures, Vol 2, The Welding Institute, Abington Cambridge, 1971, P 193
65
2-23. Low Carbon Steel: Effect of Applied Anodic Currents in 3% NaCI
.. 120 E ~
~ 100
>-t--
~
w o
80
~ 60 w
IX: IX:
~
040
20 OL-_ _- '_ _-'--------'-_'--.l-L....L--L..J'--_ _- '_ _-'--------'-_'--L-JL..l...-'LIlk<
10'
10'
CYCLES
10'
TO FAILURE
Effect of applied anodic currents on the fatigue lives of low-carbon steel in deaerated 3% NaCI solution. The corrosion rate ofthe steel in this solution is virtually zero in the absence of applied currents. Note the independence of fatigue life at currents greater than-: 40 /-LA/cm 2, the absence of an applied stress effectand the reappearance of a fatigue limit at currents less than ~ 0.2 /-LA/cm-,
The effects of salt concentration and temperature on the fatigue behavior of steels have been studied. Experiments performed on mild steel specimens in distilled water and in various concentrations of potassium chloride have shown that solutions ranging from 2 molal to 1/40 molal have virtually identical effects on corrosion-fatigue lives, but that at concentrations below 1/40 molal, the effect approaches that of distilled water, although corrosion rates increase in an almost linear manner with solution ion concentration. A similar result has been reported for deaerated 3% NaCI solution in which corrosion rates were controlled by applied anodic currents (see above chart). These observations indicate that a critical corrosion rate is a necessity to initiate corrosion-fatigue failures. Additionally, increasing over-all corrosion rates over a long range of rates has little effect on corrosion-fatigue resistance.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys." in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 344
66
2-24. Low-Carbon Steel: Effect of pH in NaCI and NaOH 60
I
I
55 /
I NORMAL SOLUTION
50
45
.,~ 40 Q
"'.H
=12.1
~ 35
UJ
II:
t-
'" 30 25 pH :10
20 .5
5
8
10
The effect of pH on the fatigue behavior of low-carbon steel in NaCI+NaOH.
The effect of stress frequency on corrosion fatigue has been studied by a number of investigators but is still not completely understood. For example, an early review of corrosion fatigue noted that it is difficult to compare the corrosion-fatigue properties of metals exposed to like environments because data reported are usually taken at different frequencies. In general, a given time was found to produce more damage at a higher frequency, but a given number of cycles was found to produce greater damage at low frequencies. For low-alloy steels in fresh water, a frequency of 1450 cycles/min produced failure in 106 cycles or II Y2 hours, but at a frequency of 5 cycles/min, failure occurred in 0.11 X 106 cycles, or 400 hours. To date, the effect of pH of aqueous solutions on corrosion-fatigue behavior has not received extensive study. A study of the effect of 0.1 N HCl on the fatigue life of steels showed greater damage in this medium than in neutral potassium chloride solutions. Tests conducted in alkaline media, at a pH above 12.1, showed that a fatigue limit is regained, this limit improving at still higher pH values (above chart). These investigators suggested that corrosion fatigue is a result of differential aeration cells, which produce pits in the metal surface, and that a high pH provides diffusion barriers (ferrous hydroxide) to oxygen on the surface. Higher fatigue limits at high pH are explained in terms of a "better and more perfect film barrier."
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys." in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 346
2-25. Low-Carbon Steel: Effect of Carburization and Decarburization
CJ)A CJ)
•
~••••
NON-DECARBED MATERIAL
~ B ...............~
t) ~
,
INTRINSIC' FOR DECARB MATERIAL
. , ...DECARBED • -COMPOSITE
LIFE Influence on fatigue SoN curve of soft surface caused by decarburization.
Parts that were made from low-carbon steel, but have high-carbon surfaces resulting from carburizing, have special microstructural factors that must be considered. From the carburizing process an intergranular oxide network may develop. This oxide may be an alloy oxide which causes alloy depletion in grain-boundary areas. As a rule, this condition is thought to detract from fatigue properties. The two exceptions may be in combination rolling and sliding contact fatigue, where the oxide network may enhance low-cycle bending fatigue-somewhat the same as does decarburization. The effect on high-cycle bending fatigue is deleterious, as is decarburization. These concepts are shown schematically in the above chart.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979,P 80
67
68
2-26. A514B Steel: Effect of Various Gaseous Environments on Fatigue Crack Propagation
r------;;;:;:;;;::::::j""". 18 16
0.4
14 2 AK = 29.7 MN/m 3/
0.3
0.2
0.1
4 2
The fatigue-crack propagation of ASTM A514B steel in various gaseous environments.
The origin of the element (such as sulfur) Onthe surface could result from its presence in the gas phase (such as for hydrogen sulfide). It could also originate as an enriched sulfur layer associated with a propagating crack, as would be the case for sulfur segregated to a grain boundary. Oxygen alone on the surface tends to drive the hydrogen-dissociation reaction rates in the opposite direction from the sulfur. The above bar chart shows how a mixture of environments can influence the fatigue-crack growth of an alloy when all the loading factors are kept constant. The main influence ofthe environment is to supply the active atoms to the vicinity of the crack tip. Subsequent interaction with the crack allows the degradation mechanism to take place. The next step in the environmental interaction is the transport of the active species to the location in the vicinity of the crack tip where the degradation mechanism takes place.
Source: H. L. Marcus. "Environmental Effects 11:Fatigue-Crack Growth in Metals and Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979. P 371
2-27. Cast 1522 and 1541 Steels: Effect of Various Surface Conditions
NONE
0.006
0.012
NONE
0.006
0.012
SHOT PEENING INTENSITY - C, ALMEN The effect of shot peening, carburization, and decarburization on the endurance ratio of normalized and tempered cast steel with cast surfaces. Plate bending fatigue .specimens were used to secure these data.
Decarburization of the surface lowers fatigue resistance. This effect, along with the beneficial effects of carburization and shot peening on the endurance ratio of cast low alloy plate specimen in bending, is shown in the above diagram. The nominally 1.2% Mn steels with 0.22% C and 0.4 I% C, respectively, were normalized and tempered to 78 and 95 ksi (538 and 65~ MPa) ultimate tensile strength, respectively. The depth of decarburization (0.05% C at the surface) was 0.06 in. (1.5 mm); that of carburization (1.15% C at the surface) was 0.08 in. (2 mm).
Source: Steel Castings Handbook. 5th Edition. Peter F. Weiser. Ed.. Steel Founders' Society of America. Rocky River OH. 1980. P 15-29
69
70
2-28. Cast A216 (Grade WCC) Steel: Fatigue Crack Growth Rate 20
40
60 80 100
200
3r---.---.--__,_~__,_,...,.,_r_--"
lTyS : 48 ksi 2
6
(331 MPo)
TEST TEMP: 7soF ( 24°C) TEST FREQUENCY: 600 cpm WOL TYPE SPECI MENS
- 4
w
UPPER SCATTER BAND ( SLOPE n : 3 ) """ 2
-.I U
r
u
-,
E 10- 6 - 8
Z
6
"0
<,
o
0::
10- 5
0::
2
8 I
6
o
0::
{.?
0::
8
~.
'
U
I
I~ 0
10-7{.?
4
~
0::
w
I
W
~
0
"0
I
4
I
I-
"0 <,
2
"0
~
I
Z
2
o. 0
6
~
u
0:: U
4
.
10- 6 '--_ _-'-_-'----'_L.-I-L.........--'-_ _---' 10 20 40 60 100 200
STRESS INTENSITY FACTOR ~K,
RANGE,
ksi ~
Fatigue crack growth rate as a function of ilK for A216 (grade weC) cast steel.
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. This is observed for a wide variety of materials and is illustrated in the above diagram for an ASTM-A2l6, Grade WCC cast steel. Some materials show a significant influence of the mean load or Klevel on fatigue crack growth rates. The ratio ofKmin to K max is used to express the mean load conditions.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, p4--16
3-1. AISI 1030 (Cast) Compared With AISI 1020 (Wrought) (\J
<, IIII
I
--r--...,
I.O....--r--..,.....--r---r--.......
'f "''0.1
ILl C
C
/WROUGHT SAE 1020 ....
,
1
~
I-
:J Q.
0.01
0:
/CAST SAE 1030
/O,W O\' '
0,'
'00"::0." 1)'0'
~
PLASTIC
",
~O
0.001
"
t; 0.0001 ~_L--..l....::-~-..l...:-....l..:~....L.:=--~ 7 100 10 102 106 10
REVERSALS TO FAILURE - 2N f Low-cycle strain-control fatigue behavior of carbon steel.
A number of techniques are available for computing the lowcycle fatigue life, although a straightforward approach is simply to compute the fatigue life from the expected cyclic plastic strain amplitude in service. Errors in computing or estimating 6.Ep produce a smaller change in the computed cyclic life than similar errors in the elastic strain range. Note that there is a large difference in slopes "c" and "b" in the above diagram. Plastic strain ranges may be computed using sophisticated finite element techniques, estimated from simple approximations such as Neuber's rule or experimentally measured in component or model tests.
Source: Steel Castings Handbook, 51hEdition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,P 4-13
71
72
3-2. AISI 1035: Effect of Gas and Salt Bath Nitriding
50
A1MQSPH£RE
NITRIDED
40
SALT BATH PRQCT_'S 0
§ lC
'0
QUENCIU:D AND TEl1PERID
AT l050 F (565 C)
~ HUKBER OF CYCLES
Torsional fatigue strength of AISI 1035 steel-stress vs number of cycles for completely reversing torsional fatigue, featuring the effects of gaseous atmosphere and salt bath nitriding on fatigue strength.
Source: J. A. Riopelle. "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals. Metals Park OH, 1977, P 286
73
3-3. AISI 1040: Cast vs Wrought
TENSILE YIELD STRENGTH STRENGTH ELONG. HARDNESS ksi MPo ksi MPo % BHN CAST WROUGHT
94 (6481 90(62~
56 (386) 56 (3B6)
25 27
187 170 350
en 50 .><:
~
,
en
J, 45
f3
g:
~
WROUGHT
}
NO
300
f3
0:
40
CAST
~
NOTCH
en
250 ~
~ 35
::::>
::::>
~
~
5(30
200 NOTCHED
X
~ 25
150 5 10
6 10
7 10
CYCLES TO FAILURE Fatigue characteristics (S-N curves) for cast and wrought 1040 steel in the normalized and tempered condition, both notched and unnotched. R. R. Moore rotating beam tests, K, = 2.2.
Cast steel suffers less degradation offatigue properties due to notches than equivalent wrought steel. When the ideallaboratory test conditions are replaced with more realistic service conditions, the cast steel shows much less notch sensitivity to variations in the values of the test parameters than wrought steel. Under the ideal laboratory test conditions and test preparation (uniform section size, polished and honed surfaces, etc.), the endurance limit of wrought steel is higher. The same fatigue characteristics as those of cast steel, however, are obtained when a notch is introduced, or when standard lathe-turned surfaces are employed in the rotating beam bending fatigue test. These effects are illustrated above.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 15-10
74
3-4. AISI 1045: Relationship of Hardness and Strain-Life Behavior
.
C\l .....
0.1
SAE 1045
or
"0
.~
Q.
E
0.01
C3: c:
>-;~-
_ _ _ 1l911HB '.:::--=----_410 ..... __ - - - 3 3 0
.
1:1 +0
(/)
0.001
- - - - 280 2211
I
10
102
103
104
6
lOll
Reversals la Failure, 2N
10
107
f
Strain-life behavior of medium-carbon steel as a function of hardness.
Strain-life curves at various hardnesses are presented in the diagram above to demonstrate the range of properties attainable by tempering. Such information, used in conjunction with life-prediction models, provides guidelines for optimizing material processing for specific situations.
Source: R. W. Landgraf. "Control of Fatigue Resistance Through Microstructure-Ferrous Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH. 1979.P 458
3-5. AISI 1141: Effect of Gas Nitriding
ATHOSI'HERE NIUIDED
ATHOSI'HERE NITRIDED niDI
QU!NCBED AND TDO'UlD
GROUND to IlDlOVE COHPOUND I.\YEII
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.
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals, Metals Park OH, 1977, P 287
75
76
3-6. Medium-Carbon Steels: Interrelationship of Hardness, Strain Life and Fatigue Life 1.0~
\.
\.
1\ \.
\\ \
0.1
1\\'\\ 600
~500~00
.
en
"
~
c;
"
t:
S '0
200 .'\. \
Hardness, HB
<; ~ t\
.~
~
,,30~
-, "'-.."
0.01
n;
.
600
s:
~ ~ 400 <, 300<, -200-
.
c: 0
-
0.001
0.0001 1
10
1M
1~ 1~ Stress reversals to failure
1~
Predicted plots of strain versus fatigue life for typical mediumcarbon steels at the hardness levels indicated above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH. 1978, P 673
77
3-7. Medium-Carbon Steel: Effect of Fillet Radii ,---
60
<-,
t--
~
--~~p~~lf:o ~
~
c
~/lf:
1ft'! d
~O'47"
on
o
r--,
~
~
gj 50
.
o x
lI
oil oil III
~
0:
li; I~
~40
I"""
z
~ o
Z
~ o
30
-
I
300,000
d
I
~(l'O'I7 d. r 'Y-. I Q ...
r- u
~S-
d.·2·/3"~ (i·Q·If;
DIcJ.:2 ~
{)r
D/d: 2
O/cl.ys
0+ ----
I II
1.000,000
10.000,000
50,000.000
I'lUMBlR OF C'{CLES ,0 FAILURE (LOG. SCALE)
~A1 UNMODIFIED
EXTERNAL 5TRE55- REl.IEVER
~&j RE-EI1TRAN, FILLET
~4 5EPARATE COLLAR
The fillet radius at a change in diameter should be made as great as possible. This cannot always be done; e.g., if the inner race of a rolling bearing must abut against a shoulder formed by the change in diameter. In such cases the stress-raising effect can be moderated very considerably by adopting one of the expedients illustrated above.
Source: G. A. Cottell, "Some Common Stress Raisers in Engineering Parts," in Failure Analysis: The British Engine Technical Reports, F. R. Hutchings and Paul Unterweiser, Eds., American Society for Metals, Metals Park OH, 1981, P 108
78
3-8. Medium-Carbon Steel: Effect of Keyway Design
5LEO-RUNNl:.R KEYWAY
PROFILED Kf..YWAY
70,000
PROFILED ...EV.......'(
o
MEDIUM-CARBON S,.EEL (NORMALISE-D)
I
o
200.000
I
I ~ I
1.000.000
10,000.000
NUMBER. DF C'(CLES,.O FI'.\LURE (LOG.StALE)
Keyways are severe stress raisers from which fatigue cracks are very liable to develop. Where bending stresses are predominant the cracks usually run transversely in the region of the keyway end, 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. Effects of various keyway designs on fatigue life are shown above.
Source: G. A. Cottell, "Some Common Stress Raisers in Engineering Parts," in Failure Analysis: The British Engine Technical Reports. F. R. Hutchings and Paul Unterweiser. Eds., American Society for Metals, Metals Park OH. 1981, p 109
3-9. Medium-Carbon Steel: Effect of Residual Stresses +600
+400 +200 NEAR-SURFACE RESIDUAL STRESS
O~-----------~-------
MPa
-200
(=
+3000 po IN/IN
• ••
TESTS STOPPED AT 10 7 CYCLES MAX.
-400
-600L---------....,...L,.-----------,J. 1.0 10 0.1 UFE CYCLES
Ie
10 6
Fatigue life relationship to near-surface residual stress.
Fully reversed fatigue tests on smooth bar specimens in medium carbon steels fully hardened show, as expected, that fatigue life increases directly with surface and near-surface residual compressive stress (see above chart). Residual stress measurements are usually made in the direction of the applied stress. The achievement of high residual compressive stress in a part requires a careful balance of the factors which affect this property and often involves a number of trade-offs which vary with the application.
Source: J. Alan Burnett. "Prediction of Stresses Generated During the Heat Treating of Case Carburized Parts," in Residual Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 44
79
80
3-10. Medium-Carbon Cast Steel: Effect of Changes in Residual Stress 25
211
.
.>:
NDR"""L.IZE ~N~
15
.4'~
I
~
:a---~
<,
-, -, ~HOT BLASi'
e;
III
i-
lJl
lil.... '4
-'
:§
A '
~
5
Ul ILl
NORH"LIZE -
<,
TEMP~
'"
"a... ~
~ ti
II
-5 111
4 CYCLES
Residual stress at completion of testing,
III
-411 .1.--+-
+--_ _--<
--+
--+-
-+-
+--'
Ill" CYCLES
Change in residual stress with cycles at constant applied stress,
The upper chart shows residual stresses existing on the completion of individual tests. The similarity to S-N curves is apparent with the exception that the curve for normalized bars (R 1 =+22 ksi) is inverted. Since the initial residual stress was known, there was a question on the manner in which the residual stress changed during the progress oftesting. To explore this point, two shot blasted bars were tested with applied stress levels of 40 and 55 ksi. The test on each specimen was interrupted periodically to measure the residual stress at that time. The results are shown in the lower chart. It is apparent that the change in residual stress is proportional to the number of cycles when the latter is represented on a logarithmic basis. The lower chart also points to the fact that the rate of change in residual stress is dependent on the level of applied stress. Since the initial and final residual stress values were known for all bars, the slope for each line could be determined.
Source: P. J. Neff. "A Quantitative Evaluation of Surface Residual Stress and Its Relation to Fatigue Performance," in Residual Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, /981, pp 127-128
83
4-1. Medium-Carbon Alloy Steels, Five Grades: Effect of Martensite Content 100
100
650
.
c,
::;;;;
600
]
"'"
.~ 550 u.
~ <,
~
500
6.
01340 .4042 _4340 05140 6. 80840
I--.I'---
'"I'--. ---
90
·Vi ~ ~.
:~
6.
-
6. n
I--.
70
All specimens 36 HRC 450
400
60 100
80
60
40
20
Martensite. %
For specimens having comparable strength levels, resistance to fatigue depends somewhat on the microstructure. A tempered martensite structure provides the highest fatigue limit. However, if the structure as-quenched is not fully martensitic, the fatigue limit will be lower (see graph above). Pearlitic structures, particularly those with coarse pearlite, have poor resistance to fatigue. S- N curves for pearlitic and spheroidized structures in a eutectoid steel are provided in chart 4-40 (p 122).
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 676
84
4-2. Medium-Carbon Alloy Steels, Six Grades: Hardness vs Endurance Limit H-ll 1----1I---t---I----1t-"'1 Aus t ernpered
160 150 140 Vi a. 0 0 0
130
E --'
100
Q)
120 110
90
u
c ~
• - SAE SAE .a._ SAE 0 - SAE G - SAE Q - SAE
:J "0
6-
c w
50 20
30 Rockwell
40
"c"
4063 5150 4052 4140 4340 2340
50
60
Hordness
Relation of hardness and fatigue strength for several steels.
The above chart and other data can be used to show the importance of limiting the system to low and intermediate hardnesses as well as to point out the importance of residual stress in fatigue. These data from Garwood, Zurburg and Erickson show a very tight linear relation up to about 40 HRC. Above that hardness, the relation deviates from linearity, seemingly depending on carbon content. Carbon, however, is in an intermediate role here, because it affects temperability. Because response to tempering is dependent on carbon and alloy levels, it was necessary for samples of different grades to be tempered at different temperatures to achieve the same hardness; consequently, a variety of residual-stress conditions resulted. The tempering temperatures were necessarily sufficiently high to obtain 40 HRC; the residual stresses were reduced to a very low level, making all samples similar in that usually the tensile strength for small sections decreases with increasing section size and I or decreasing hardenability to compressive values. The sequence of transformation from surface to center, together with the temperature gradients, governs the outcome.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 73
4-3. Medium-Carbon Alloy Steels: Effect of Specimen Orientation
Steel
BOO 100
s:
::;:
.~
.>!
600 BO
.
5,
.~
~
.
5, 60 "g
400
u,
u,
40 200
4140 Hardness, HRC 30
X4340 32
4027 44
4063 46
4032 4B
Avg tensile No. of strength Hardness, tests(a) MPa ksl HRC
Longitudinal 4027 4063 4032
Tests 11 1179 171 12 1682244 11 1627236
Transverse Tests 4027 10 4063 9 4032 10
37 to 39 47 to 48 46 to 48
1130 164 34 to 39.5 1682 244 47 to 48.5 1254 182 47.5 to 48.5
(a) Number of fatigue specimens. For 4140 steel, 50 longitudinal and 50 transverse specimens were tested; for 4340 steel, 10 longitudinal and 10 transverse.
It must always be considered that in rolled steels fatigue behavior is affected significantly by specimen orientation. Shown above is the effect of orientation relative to fiber axis resulting from hot working on the fatigue limit of low-alloy steels. Through hardened and tempered specimens, 6.3 mm (0.250 in.) in diameter, were taken from production billets. Specimens for each grade were from the same heat of steel, but the tensile and fatigue specimens were heat trated separately, thus accounting for one discrepancy in hardness readings between the chart and the tabulation above. Fatigue limit is for 100 million cycles.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 677
85
86
4-4. 4027 Steel: Carburized vs Uncarburized
0.02 (\J
.....
0.01
Bending Fatigue
• 4027 Carburlzed b. O.OOS"Ca.. D O.OIS" Ca.e o 0.03S"Ca..
"0
~
0.005
ji
E
~
c
.~
.. 0.002
Ul
0.001 L...-_--'-_----'--;.----'-....-_'-;;-_........_--'-;;_--' 102 103 104 lOS 10 1 Reversals 10 Follure,2N f Bending-fatigue results for uncarburized and carburized 4027 steel.
Bending-fatigue results supporting the validity of the effect of carburizing are presented in the above curves. An uncarburized baseline curve is compared with curves for three case thicknesses. As predicted, all carburized specimens show inferior low-cycle resistance. At longer lives, the thinnest case offers some improvement but tends toward the baseline as a result of subsurface failure initiation. The thickest case, which shows the greatest life improvement and has been found to exhibit surface failure initiation, seems close to optimum.
Source: R. W. Landgraf, "Control of Fatigue Resistance Through Microstructure-Ferrous Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OR, 1979, P 463
4-5. 4120 Steel: Effect of Surface Treatment in Hydrogen Environment
200°6 1600
600
25 Cr '"'A 4 - annealed .4~.= 1.2%. H1 (10MPo) £ =aOO3 sec:' Cl bose metal - mech. pot. m bose metot - chem. pol.
prior to surface treatmMt
surtoce treatments
Effect of surface treatment on fatigue life in hydrogen environment for a O.23C-O.98Cr-O.22Mo steel.
In the above bar chart, the effect of surface treatment on fatigue life is summarized. The base metal was mechanically polished before surface treatment. For comparison, pot galvanizing and Ni-plating have been performed after mechanical as well as after chemical polishing of the base metal. The results after chemical polishing are given above in the form of dashed columns. The galvanizing such as Ni- and ZN-plating is by no means an appropriate method to increase the fatigue life in hydrogen in spite of the reduced surface roughness and protecting effect. This is because the galvanizing produces a relatively high tensile residual stress and the deposits possess generally poor ductility.
Source: Kyong-Tschong Rie and Werner Kohler, "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High Pressure Hydrogen Environment," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Presouyre, Eds., American Society for Metals, Metals Park OH. 1982, P 380
87
88
4-6. 4120 Steel: Effect of Surface Treatment in Hydrogen Environment 500'.-----~---~-----,-------,
,
hammered
I / alvanized I
e-
300
III
~
I Ia/hed
I
III
200
I
Ni- la/ed
-
-1 '25 erMa, -annealed
number af cycles N
Effect of surface treatment on cyclic now curve in hydrogen environment.
The above graph shows the cyclic strain hardening and softening curves for different surface treatments. It can be seen that the fatigue behavior in hydrogen environment can be improved by some surface treatments.
Source: Kyong-Tschong Rie and Werner Kohler, "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High Pressure Hydrogen Environment," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Pressouyre, Eds., American Society for Metals, Metals Park OH, 1982, P 379
4-7. 4120 Steel: Effect of Various Surface Treatments on Fatigue Characteristics in Air vs Hydrogen
..e
..
2.0
..
0.
,
''';'.
"'l
1.0
.....
,:~
.
25 erMa' -annealed -+~ i: =0.003 sec-' -H1,IOMPo) S --air
g ~
s
0•• Ni-ptatea
0,2 60
0,+ mech. 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. Steel contained 0.23 C, 0.98 Cr and 0.22 Mo (4120).
Source: Kyong-Tschong Rie and Werner Kohler, "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High Pressure Hydrogen Environment," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Pressouyre, Eds., American Society for Metals, Metals Park OH, 1982, p 380
89
90
4-8. 4130 Steel: Fatigue Crack Growth Rate vs Temperature in Hydrogen 10- 3
I
0
O"ty = 1330 MN m- 2 I O"tu = 1660 MN m- 2
o
O"ty = 1190 MN m- 2
I
O"tu =1310 MN m-
2
K =40 MN m- 3 / 2
o
.,
U
E
I
3.4
3.8
4.2
I 4.6
5.0
liT, oK-I
Crack growth rate versus temperature in hydrogen gas, 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. Although the diffusivity of hydrogen does vary significantly among metals and alloys, it is nevertheless always several orders of magnitude larger than the diffusivities of other species. Consequently, hydrogen transport is a prominent feature of discussions of hydrogen-induced crack growth kinetics, and ofthe unique strain rate and temperature dependence of hydrogen embrittlement. Nelson and Williams reported the first complete investigation ofthe kinetics of crack growth in high strength steel exposed to hydrogen gas (see graph above).
Source: Herbert H. Johnson, "Keynote Lecture: Overview on Hydrogen Degradation Phenomena," in Hydrogen Embrittlement and Stress Corrosion Cracking, R. Gibala and R. F. Hehemann, Eds., American Society for Metals, Metals Park OH, 1984, P 17
4-9. 4135 and 4140 Steels: Cast vs Wrought 85 80
UNNOTCHED NOTCHED
75
65
t:
I (f) (f)
60
I-
0
D~-& °o~ ............... o
"<, 80--0~ ~" 06.., ....
o
0--
6,6..
100
80 0
.....
o,
""8<,6 . . - 6 .__
55
~
O-B~
UJ
a:
••
0
4140 (LONGITUDIAL) }NO.32 4140 (TRANSVERSE)
70
A
6.
4135 CAST STEEL
(f) (f)
•••
50
\~\A
(f)
45 40
UJ
a:
I60
\\.\.
.
,,~& .........
35
~~
30 25 104
(f)
·-A-A_
~.--
'.~-.-
105
106
40
107
CYCLES TO FAILURE SoN curves of a normalized and tempered AISI 4140
wrought steel in the longitudinal and transverse direction and cast 4135 steel normalized and tempered. Tensile strength for wrought steel: longitudinal, 110.0 ksi (758 MPa); transverse, 110.7 ksi (763 MPa); cast steel: 112.7 ksi (770 MPa).
In general, if the longitudinal and transverse ductility, impact, or fatigue property values of rolled steel are averaged, they will be about the same as properties of cast steel. One example of this is shown in the S-N curves presented above. For these, a 4140 rolled steel was tested in fatigue in the longitudinal and transverse position and compared with a similar Cr-Mo cast steel.
Source: Steel Castings Handbook, 5th Edition. Peter F. Weiser, Ed.. Steel Founders' Society of America, Rocky River OH, 1980. P 3-16
91
92
4-10. 4135 and 4140 Steels: Cast vs Wrought STRENGTH YIELD TENSILE
ELONG % 113 (779) B7 (560) 43 110 (758) 80 (552) 61 III (7651 81 (558) 30
ksi
CAST 4135 WROUGHT 4140-L -T 80
WROUGHT
75
CAST
(MPo)
ksi (MPol
BHN 223 217 217
_ 500
~
70
I
b
65
W o ~
60
...........
' ......
fh . . . . . . . . .
UNNOTCHED'
.
-,
LONGITUDINAL
------~
<, <,
400 11.
---
en en
~
«
en en w
lr
I-
en
\ \
45
\
\
30
I
w
lr 300 I-
en
""
I ',
40 35
~
,1__ -.
55
::i
11. 50
o
,~RANSVERSE
NOTCHED LONGITUDINAL ''-...... AND TRANSVERSE - ---NO FAILURE
=*
200
106
CYCLES
TO FAI LURE
Fatigue characteristics (S-N curves) for cast and wrought 4100 series steels, quenched and tempered to the same hardness, both notched and unnotched.
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. The fatigue life increases as the cyclic stress amplitude decreases. For ferrous alloys a true endurance or fatigue limit is reached below which fatigue failure is not observed. The data presented in the above S-N curves illustrate several important points. First, a fatigue limit is evident. That is, below a certain cyclic stress amplitude, fatigue failure will not occur for any arbitrarily large number of cycles. Secondly, while the fatigue properties of cast steel are lower than those obtained with the wrought steel, it has less anisotropy. And, finally, the presence of a notch equalizes the fatigue properties of cast and wrought steels. 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.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 4-8
4-11. 4140, 4053 and 4063 Steels: Effect of Carbon Content and Hardness 1000,-------------,--------,--------.--------; 140
o to 2 micro-in.finish 900.1--------/-------t--------t
130 120
~
8001--------/-------t--,
·in
Q.
sc
::;;
110
~.
·E ~
.~
~
7001-------j-----;
100
:J
.~
. :J
.~ u,
U.
90
6001-------:
80 70 60 30
40
50
60
70
Hardness, HAC
Effect of hardness and carbon level on fatigue limit of alloy steels.
As shown above, when steels are hardened to 45 HRC or higher an increase in carbon content can increase fatigue limit. Although other alloying elements may be required in order to attain desired hardenability, they have little effect on fatigue behavior.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, p 676
93
94
4-12. 4140 Steel: Effect of Direction on Fatigue Crack Propagation 10-'
10'
T
L
..... Q)
~
U
..... Q)
10-'
~
U
>U
10:'
>U
.....
.....
E E
E
....
....E
C121(j'
cI2 Ill"' -
'tJ'tJ
'tJ'tJ
10·
1-1-_ _' - - _ - - - ' - _ - ' - - ' - - ' - - _ - - - - ' - '
10
20
304050
AK (MPaV'ffi) a
100
10·
L.L_ _-'--_---J'---'--'--'-_--,-lJ
100
5
AK(MPaV'ffi) b
Fatigue crack propagation in an AISI 4140 steel: (a) longitudinal direction (parallel to rolling direction); (b) transverse direction (perpendicular to rolling direction).
The Paris power law, which describes the crack propagation rate in stage II for a series of metals, is very useful because of its extreme simplicity. For example, it has been observed experimentally that data points in the form of log (do / dN) versus log ~K for a given material (constant metallurgical structure) from three different samples-edge crack in a compact tension sample, through-thickness central crack in a plate, and plate containing a partially through-thickness crack -all fall on the same line. Also, there is experimental evidence that shows that the stress level by itself does not influence the fatigue crack growth rate for stress levels below the general yielding. Thus, it can be considered that the parameter ~K describes uniquely the crack growth rates for many engineering applications. However, the structure of material can influence fatigue crack growth rates drastically; the value of m can change a lot. The above charts illustrate the directionality in the fatigue crack propagation rate in an AISI 4140 steel. The exponent m has a much higher value in the transverse direction than in the longitudinal (rolling) direction, due to the presence of elongated inclusions.
Source: Marc Andre Meyers and Krishan Kumar Chawla, Mechanical Metallurgy: Principles and Applications, Prentice-Hall, Inc., Englewood Cliffs NJ, 1984, p 714
4-13. 4140 Steel: Effect of Cathodic Polarization
Carras.....
LLi-O.4 I
Potential . - . _ " - - -
.~
•
(fj
•
(/) -0.6 ~
o
>-0.8
.r <{
~ -1.0
w
I-
o
0...
•
•
-1.2
10 5 106 CYCLES TO FAILURE The effect of cathodic polarization on the fatigue behavior of 4140 steel (heat treated to HRC 52) in 3% NaCI solution at a stress level below the fatigue limit in air is shown above. The use of cathodic protection to prevent corrosion fatigue of steels depends sensitively on the hardness of the steel. For example, cathodic protection of a 4140 steel was shown to be feasible for hardness values of Rockwell C 40. At higher hardness values, an improvement in fatigue resistance is observed for moderate cathodic potentials, but complete protection is not possible. At potentials large enough to inhibit corrosion fatigue for softer steels, a decrease in fatigue resistance is observed, presumably due to hydrogen embrittlement (note above chart).
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OR, 1979, P 360
95
96
4-14. Cast 4330 Steel: Effects of Various Surface Conditions SURFACE ROUGHNESS- RMS - rnrn- 10- 2 .5
2.5
2
1.5
3
FUL L Y MACHINED
o
/
f===:J~ ,
.... 0.2
L,\ ~ w
SILICA PRIME INV. ZIRCON INV. SILICA ALUMINA C 6 C 16 WASH ZIRCON GR.S.
INV CRYOLITE
HIGH
WASH
""""';;:::::====~~:.:;:=~:::..CHROM ITE
GR. S. PROPR. WASH _ _ _ _==--========,JS~IL~ICAGR.
COPE
a
DRAG
200
MACHINED
400
600
800
1000
SURFACE ROUGHNESS - RMS- 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)]. Fully reversed plate bending tests. .
Plate bending tests for quenched and tempered low alloy cast 4330 steel indicate that investment cast surfaces, or conventional castings produced with special mold washes, performed equal to, or better than, fully machined and polished plate specimens. The data also suggest a tapering off of the surface effects on the endurance ratio at 600 or more RMS surface roughness as indicated in the above diagram.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America. Rocky River OH, 1980, P 15-29
4-15. 4340 Steel: Scatter of Fatigue Limit Data Tensile strength, ksi
12 800 iir=0-
-
-
-
160 ---'-T-'------
-
-
200 ----,rr--
-
-
-
240 rr-r --
-
-
280 .,---, 110
7001-------1---------11--------1--------:::J 100 50% survival
e:;;
90
Ii 600 f-------t--------::;;;Io--"""'-----t------=90%
~ c '" .~
.""'"
~
1;: 99%
500
1----.....~l=____=,....",~~~k:;;o>'-=::::::=:1f===-----l
~
80 c '" .~
70
~
E
~
~
400 f - - - - - - - - ' - - - t - - - - - + - - - - - - t - - - - - - - i 60 Approximately 1000 specimens. 1 heat 50 300 '--
800
L-
1100
-----''-----
~~
_:_:_'
2000
Interrelationships of alternating stress, tensile strength and expected percent survival for heat treated 4340 steel.
These data show survival after I a million cycles of AISI-SAE 4340 steel with tensile strengths of 995, 1320, and 1840 MPa (144, 191, and 267 ksi). Rotating-beam fatigue specimens tested at 10 000 to II 000 rpm. Coefficients of variation range from 0.17 to 0.20. From these data it is evident that scatter increases as strength level is increased.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 678
97
98
4-16. 4340 Steel: Strength vs Fatigue Life
'.
0', = 0, = 174 ksi / ' F..tigue strength eOr'ie;ent 0. - 0',{2N,)b - 17412N,ro.o9
/
I Fatigue strength
exponent
=
slope
/ =
b
=
-0.09
-
Reversals to failure. 2N f
Typical data for strength versus fatigue life for annealed 4340 steel.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 672
4-17. 4340 Steel: Total Strain vs Fatigue Life 100, - - - -....- - - - - - , , - - - -....------,-----r-------,------,
~
10-11----+---"O:-"'.--I-----r- 2
= l1€p +
2
~ E
=
0 5Bl2N r O.57 + 0 0062(2N r O.09 . , . ,
10-2 1 - - - - - + - - - - 1 - - - - - - - 1 " - " ".......:----1----+-----+------1
10-4l::-
--'-:-
---lL::-
-'-::
-'--:--
-'-::--
-..L:--~"___J
100
Typical data for total strain versus fatigue life for annealed 4340 steel.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 672
99
100
4-18. 4340 Steel: Stress Amplitude vs Number of Reversals
4 10
f
( V = Uf =1200 MPa C Q.
3 10
~ C
b ~
0
2
~IN
10
1
10
Stress amplitude (!:J.aj2) versus number of reversals (2 N J) for AISI 4340 steel.
It is convenient to consider separately the elastic and the plastic components of strain. The elastic component can be readily described by means of a relation between the true stress amplitude and the number of reversals (i.e., twice the number of cycles):
~E
a (a'J) E
_e_=_o_= _ _ (2N)b
2
E
f
where ~Ee12 is the elastic strain amplitude, a o the true stress amplitude, aj-the fatigue strength coefficient (equal to stress intercept at 2NJ = 1), NJthe number of cycles to failure, and b the fatigue strength exponent. This relation is an empirical representation of the S- N curve above the fatigue limit. The above chart shows an application of this relation to SAE 4340 steel. It was observed that fatigue life increased with decreasing b. Morrow, based on energy considerations, showed that the fatigue strength exponent is given by: n' b=--1 + 5n' where n' is the cyclic hardening coefficient. Thus, the fatigue life under elastic cyclic conditions (whether stress- or strain-controlled) increases with a reduction in n'. Of course, the higher the material coefficient aj, the better it is for fatigue. There is evidence that ajis approximately equal to aJ' the monotonic fracture strength. The plastic strain component is better described by the Manson-Coffin relation: ~E
-p-=
2
r
Ej( 2NJ
where ~Ep12 is the plastic strain amplitude, Ej is the ductility coefficient in fatigue and is equal to strain intercept at 2NJ = 1, 2NJ is the number of reversals to failure, and c is the ductility exponent in fatigue.
Source: Marc Andre Meyers and Krishan Kumar Chawla, Mechanical Metallurgy: Principles and Applications. Prentice-Hall, Inc., Englewood Cliffs NJ, 1984. p 697
101
4-19. 4340 Steel: Effect of Periodic Overstrain 2000 250 1500
e
200
::;;
"c '" ~ ~
e
.>l
1200
1J, c
1000
Vi
o No overstrain or single over-
0
strain at beginning of test
150 0
~ ~
e
0
Vi
• Periodicoverstrain
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. Shown above is the effect of periodic large strain cycles on fatigue life of AISI-SAE 4340 steel hardened and tempered to a yield strength of 1100 MPa (160 ksi).
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH. 1978, P 681
102
4-20. 4340 Steel: Estimation of Constant Life Minimum SIren, ksi
600
--400
~200
200
.00
600
800
1000
1200
1'00
Minimum Sites'. MPa
Potter has described a method for approximating a constant-lifetime fatigue diagram for unnotched specimens. Using this method, a series of points corresponding to different lifetimes are calculated and plotted along the diagonal line on the left side (R = -1). Each of these points is connected by a straight line to the point of the other diagonal (R = 1.0) that corresponds to the ultimate tensile strength. A comparison between the estimated constant-lifetime diagram and the experimentally determined diagram is given in the above illustration. Here is presented a comparison between a calculated constant-life fatigue diagram (solid lines) and experimentally determined data (dashed lines). The calculated lines correspond well with the experimental lines. Generally, the predicted lines represent lower stresses than the actual data. Estimating fatigue parameters from the Brinell hardness number provides more conservative estimates. These results are only approximations, and the methods may not apply for every material.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 681
103
4-21. 4340 Steel: Effect of Strength Level on Constant- Life Behavior Minimum stress. ksi
1.0 200
.
c,
:;;;
~ E
150
1000
"'" ~
800
E
"E
'x
~
E
"1""""')
600
100
O't;..; :9~
600
"""~~'1;1.()
..
400
10 6 cycles lifetime
50
200
o'----_---'-__---'--__--'--__-'---__ _ __''"____ ~
-1200
-1000
"'1100
-tiOO
-400
-200
____<_ ____L_ ____'___ _ ___'____ _ _~
200
400
600
800
1000
L __
1200
_ _ '_ _____'
1400
0
1600
Minimumstress, MPa
Constant-lifetime fatigue diagram for AISI-SAE 4340 alloy steel (bar), hardened and tempered to tensile strength levels of 860 MPa (125 ksi), 1035 MPa (150 ksi), 1380 MPa (200 ksi) and 1790 MPa (260 ksi), All lines represent fatigue lifetimes of one million cycles.
It may be noted that lives of the specimens at the three higher strength levels are about the same; the scatter in data is at least as great as any real differences in fatigue life among specimens.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Jronsand Steels, American Society for Metals, Metals Park OH, 1978, P 669
'x
~
104
4-22. 4340 Steel: Notched vs Unnotched Specimens Minimum stress,ksl
o
160
~
~ 800
~ ,
1;;
E E
"~
100
~
E
~ "w
600
:!E
~
400 60
200
O'--_ _....L-1000
-een
' -_ _---'-
-il00
-400
...L-_ _- ' " ' " -_ _--'---_ _- - ' ' - -_ _....L-
-200
200
400
600
-'--_ _- ' -
800
1000
-'0
- ' -_ _
1200
1400
Minimum stress.MPa
Constant-lifetime fatigue diagram for AISI-SAE 4340alloy steel (bar), hardened and tempered to a tensile strength of 1035 MPa (150 ksi), Solid lines represent data obtained from unnotched specimens; dashed lines represent data from specimens having notches with K, = 3.3.
Source: Metals Handbook, 9th Edition. Volume I, Properties and Selection: Irons and Steels, American Society [or Metals, Metals Park OH. 1978, p 667
4-23. 4340 Steel: Effect of Decarburization 1800 250 1500 200
0
...
Ie. If. 1200 :;:
~ ;
c
900
.~
~
~ 600
0
0 0
••
o Not decarburized
0
•
• Decarburized
0 0
••
150"" ::i e' :;; 0> C .~
0 0
100 ~
r.
~
.....
~ 300
-
50 >-
Number of cycles to failure
Decarburization is the removal of carbon from the surface of a steel part; as indicated in the above S-N curve, it significantly reduces the fatigue limits of steel. Decarburization offrom 0.08 to 0.76mm (0.003 to 0.030 in.) 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. When subjected to the same heat treatment as the core of the part, the decarburized surface layer is weaker and therefore less resistant to fatigue than the core. Hardening a part with a decarburized surface can also introduce residual tensile stresses, which reduce the fatigue limit of the material. 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).
Source: Metals Handbook, 9th Edition. Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 674
105
106
4-24. 4340H Steel: Effect of Inclusion Size 1100
1000
--.r--..
s::;; ~ 900 1;: en
c
.~
E 800
s
;;:
<,
o Small inclusions
• Large inclusions
<,
~
140
~
~ ~.
-
700
1
100
Number of cycles to failure
Fatigue life of two lots of AISI-SAE 4340H steel; one lot (lower curve) contained abnormally large inclusions; the other lot (upper curve) contained small inclusions.
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. Large spherical inclusions, about 0.13 mm (0.005 in.) in diameter, were observed in the fracture surfaces of these specimens; the inclusions were identified as corundum or silicate particles. No spherical inclusions larger than 0.02 mm (0.00075 in.) were detected in the other specimens. Large nonmetallic inclusions can often be detected by nondestructive inspection; steels can be selected on the basis of such inspection. Vacuum melting, which reduces the number and size of nonmetallic inclusions, increases the fatigue limit of 4340.
Source: Metals Handbook, 9th Edition, Volume 1, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 673
4-25. 4340 Steel: Influence of Inclusion Size Xl0
3
200 190 180 STRESS RANGE. Ib/on 2
• ••
170
tal
160 150 140
•
130 i20 103
" " ,;: :I
•
SURFACE INCLUSIONS
a SUBSURFACE INCLUSIONS
DIAMETER, on
0001-
I--~~ CYCLES
SoN curve and dependence oflife on inclusion size for AISI 4340 steel.
Typical initiated crack sizes are l-lO Mm. As this is an order-of-magnitude greater than dislocation substructure sizes, such an initiated crack will behave as in a continuum. For materials with lower stacking fault energy cross-slip and PSB formation is difficult, thus inhibiting initiation. For such materials crack initiation can occupy a significant fraction of life. Other microstructural sites for initiation are discontinuities such as grain and twin boundaries, the latter being particularly operative in hcp metals. Usually, however, at ambient temperatures it is the dislocation substructures which dominate initiation. For stronger, more complex alloys planar slip behavior dominates, making localized slip bands the initiation sites cfthe random notch-peak topography generated by shearing a pack of cards. The interaction of slip bands with second-phase particles (inclusions, precipitates) can produce a local stress concentration which cracks the interface, producing a surface crack. The above SoN curves show the results of this process for a low alloy steel. Variations in fatigue life relate to variations in inclusion size. As well as debonding, oxide or carbide particles can crack under concentrated localized stresses.
Source: B. Tomkins, "Fatigue: Mechanisms," in Creep and Fatigue in High Temperature Alloys, J. Bressers, Ed., Applied Science Publishers Ltd, London, England, 1981, p 115
107
108
4-26. 4340 Steel: Effect of Hydrogenation; Static Fatigue Normal notch strength
=
Ai>
.~, 275 250
'~
225
-. "\\ \
'00
.x:
ul (/) ~
200
(;)
"0
.~
a. C«
175
e\
125
\ \
100
:\
150
~
1
--
"..
Uncharged ++-
+
\
Bake 24 hr -
'---
0\ \
"\
\ \.
+
<,
Bake 18 hr
1
\ t-,
~
75
+
I'\.. '"
300 ksi
Bake 12 hr
Bake 7 hr
--
Bake 3 hr
•
Bake 0.5 hr
-.-..
50 0.01
0.1
1 10 Fracture lime, hours
100
1000
Static fatigue curves for quenched and tempered 4340 notched specimens charged with hydrogen and baked at 150°C (300 OF) for the times shown.
There are many embrittling effects of hydrogen on steels: the ultimate strength of a steel may be reduced; ductility as measured by total elongation to fracture or reduction of area may be decreased; and crack growth may be significantly accelerated. The hydrogen responsible for these effects may be present in the environment external to the steel or may be present internally as a reslt of steelmaking or processing operations such as pickling or electroplating. Hydrogen may promote a transition from a ductile to brittle fracture mode or it may reduce ductility without a change in fracture mode. The graph above shows the effects of baking at 150°C (300 OF) on the static fatigue (sustained loading) of the hydrogen-charged specimens. Increasing baking time effectively lowers hydrogen content even in the plated specimens, and sufficient baking eventually restores the strength of charged specimens to that of uncharged specimens. The horizontal portions of the curves in the graph above are designated as static fatigue or endurance limits, i.e., the stress level below which failure would not occur no matter what the duration of stress application. As hydrogen content is decreased by baking, the static fatigue limit increases. The specimens used to obtain the above data were notched and therefore the static fatigue limits hold for that particular notch geometry. In general, the sharper the notch, the lower the static fatigue limits, an indication that a critical combination of hydrogen concentration and triaxial stress state is required for crack initiation.
Source: George Krauss, Principles of Heat Treatment of Steel, American Society for Metals, Metals Park OH, 1980, p 223
4-27. 4340 Steel: Effect of Hydrogen
400
4340 250,000 PSI
350
~
300
H
250
co P-t
0 0 0
200
~
•
......
.....
......
r-I
Cf) Cf)
~
co
150 100
RATE UNEMB. EMBRIT. PLATED IN LAB 1000 RPM COMMERCIALLY PLATED 50 250 RPM 0 PLATED IN LAB 200 RPM 'V PLATED IN LAB .33 RPM 0 10,000 1000 10 100 CYCLES TO FAILURE
<>
•• •.
Schematic representation ofthe effect of cycling rate on theS-N curve of hydrogen-containing 4340 steel, heat treated to a strength level of 250,000 psi.
Source: George Sachs, "Test Methods for Evaluating Hydrogen Ernbrittlement," in Materials Evaluation in Relation to Component Behavior (Proceedings of the Third Sagamore Ordnance Materials Research Conference). Syracuse University, Syracuse NY, 1956, P 508
109
110
4-28. 4340 Steel: Effect of Nitriding
----- -- ---
ATMOSPHERE HITfUDED
------------
ATMOSPHERE NITRJDED TJfEN
GROUND TO RDlOVE COMPOUND
lAYER
5'>
QUENCHED AND TD1PERED AT 1050 f (565 C)
HUHBER OF CYCLES
10 ~
10 5
S-N curves for 4340 steel, gaseous atmosphere 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, Metals Park OH, 1977, P 287
4-29. 4340 Steel: Effect of Nitriding and Shot Peening 1200 r-----,------,-----,,---,---,,------,------,-----------, 160 Nitridedcrankshafts
120 800 100
:Ii
~ 6001----+----' Normally heat treated
500 I----+-crankshafts --+--+--+----+----+-----,---------'5,....---1
350 L -_ _--'-
- ' - _ - - - ' _ - - ' - - - - ' ' - -_ _---...
----'_.L.>...
----'
105
Cvcles to failure
Comparison between fatigue limits of crankshafts (S-Nbands) and fatigue limits for separate test bars, which are indicated by plotted points at right.
Mechanical working of the surface of a steel part effectively increases the resistance to fatigue. Shot peening and skin rolling are two methods for developing compressive residual stresses at the surface of the part. The improvement in fatigue life of a crankshaft that results from shot peening is illustrated in the above curves. Shot peening is useful in recovering the fatigue resistance lost through decarburization of the surface; decarburized specimens were shot peened, raising the fatigue limit from 275 MPa (40 ksi) after decarburizing to 655 MPa (95 ksi) after shot peening.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 674
111
112
4-30. 4340 Steel: Effect of Induction Hardening and Nitriding 160 ._ 150 \ a.. '"
:5 140 C!.
~-130
e u;
120
110
'r-.
\
\
As ni ide
\ \.
\ "I'
ilrid ~
"1\"'r--.
nducti n har en d
DU
Indu tio a dened
10 5
10 6
10 7
Cycles to Failure As demonstrated in the above S-N curves, fatigue tests of AISI 4340 steel in various surface hardened conditions show that combined treatments produce endurance limits between those developed by separate treatments.
Source: Sander A. Levy, Kenneth E. Barnes and Joseph F. Libsch, "Combining Nitriding With Induction Heating Pays a Bonus," in Source Book on Nitriding, American Society for Metals, Metals Park OH, 1977, P 241
4-31. 4340 Steel: Effect of Surface Coatings 190r---------------------------,
0---
180
_
----
O~---------<>6
170
~
Cr + SFl
<>-
~
Cd+
O .. ~
~Chromate
~1f8are
~160
j ISO
o R • 0.8
140
2000 cpm RT
Cr +
~TUngsten Carbide SFl
we - Shot Peened
+
lJOL-----,---,-.w..u.uJ.,-----,---,-.w..u.u.l.;-----'---'-.L..LlllJJ.,-----'---'-.L..Lu.u-,;---'----'-L.LJL..LUJ
SoN curves (axial tension) of bare and coated 4340 steel in air environment.
Air Stress
3.5% NaC1
Test
Condi tion
ksi
MN/m 2
Change, %
Rotating Bending R = -1
Bare Cd + Chromate Cr + Ory Film' WC + Dry Film' WC + Cori cone + Dry Film'
105 105 95 90
so
724 724 655 621 621
0 -9.5 -14.3 -14.3
Bare Cd + Chromate Cr + SFL'
160 165 175
1103 1138 1207
WC + SFL'
140
WC + Cori cone + SFL'
140
Axial Tension R = 0.8
Stress MN/m 2
Change, %
90
13B 552 621 621 621
-81 -24 -14.3 -14.3 -14.3
+3.1 +9.4
110 165 90
758 1138 621
965
-12.5
60
414
965
-12.5
60
4.4
ks; 20 80
so so
-31. 2 0 -43.8t -4B. 6~ -62.5t
-57~
-62.5 t -57f
'Shot peened tCompared to bare alloy air value ~Compared to coated alloy air value
The above SoN curves, in conjunction with the table, contain the data obtained in air for 4340 steel, bare and coated. Fatigue data at 107 cycles showed that the cadmium and chromium electroplates, particularly the chromium, improved the fatigue strength. They were similar in both rotating bending and axial tension fatigue tests. But in NaCl solution, significantly greater reductions in axial fatigue strength of the coated alloys were observed due to environmental effects, which remains to be elucidated. Since the Cr and WC hard (brittle) coatings have a relatively low intrinsic fatigue strength in comparison with the steel, they will become discontinuous at a relatively low stress level owing to the development of fatigue cracks. (The Cr normally contains internal cracks.) These cracks will permit access ofthe corrosive NaCI solution to the steel base at the root of the fatigue crack. In the case of the axial tension test (high steady tensile load), it may be easier for the environment to reach the crack tip.
Source; M. Levy and C. E. Swindlehurst. Jr., "Corrosion Fatigue Behavior of Coated 4340 Steel for Blade Retention Bolts of the AH-I Helicopter.Yin Risk and Failure Analysis for Improved Performance and Reliability.John J, Burke and Volker Weiss. Eds .• Plenum Press. New York NY. 1980. P 275
113
114
4-32. 4340 Steel: Effect of Temperature on Constant-Lifetime Behavior Minimumstress, ksi
150
'"
c,
:2
";;;
'"
800
e 1;; E E
"
"x
100 600
"1'/.
~~.,
~
E :J E
"x
"00
~
107 cycles lifetime
Q)lj...;
".9s
400
~
..
~~
J'J'
..
'1"..0
6'a 1:>
50
200
0"--_ _-'----_ _--'--_ _---'--_ _---'_ _--'''"---_ _--'---_ _---'--_ _----' -800 600 -1000 -200 200 400 --1l00 -400 Minimum stress, MPa
'"-_ _--'---_ _---'--_ _----'0 800 1000 1200 1400
Constant-lifetime fatigue diagram for AISI-SAE 4340 alloy steel (bars) hardened and. tempered to a tensile strength of 1035 MPa (150 ksi) and tested at the indicated temperatures. Solid lines represent data obtained from unnotched specimens; dashed lines represent data from specimens having notches withK1 = 3.3. All lines represent lifetimes of ten million cycles.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 669
4-33. 4520H Steel: Effect of Type of Quench
z .... I
p:\ ....:l 0 0 0
40 30 20
......
~
....:l
l=l
~
~
~4",~...
DIRECf (OIL) QUENCH (COLD OIL) /
~.;~~.'::......
/
~....
~1ARQUENCH /-.:::••••••••••••
10 8 6
(400 0F OIL)'
-........:.::-~ •••••••••••••
5
-----.....:..... ---
4 3 2 10
10 3
10 4
CYCLES - N Effect of quench type on fatigue of carburized differential cross.
Since 4520H steel is relatively low in hardenability for the part involved, the depth-hardening characteristics of the two groups were significantly different. The marquenched group had shallower case depths, which resulted in fracture origins below the surface at the case-core junctures. However, when the comparison was made with higher-hardenability steels, with sufficient gradient strengths and thus all fracture origins at the surface, the difference was very slight, though still in favor of direct quenching. This is consistent with what is known concerning the differences in residual stress, which in this case would have been the only other contributing factor. In other instances-such as for gears, where distortion could be a factor-the results might turn out differently for the marquenching.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 92
115
116
4-34. 4520H Steel: Effect of Shot Peening
40 ~,
30
CQ
,...J
20
0 0 0
......,
§ ,...J
c: u..; if;
l:>:
F
~
~':".':'.":
.
~.;.......
---~
10
SHOT PEE~D
:::.:.:
/
.
lJNPEE\'ED~ __••••••••••••
8
----...:.... --
6 5 4 3
2
10
10 2 CYCLES - r\
Effect of shot peening on fatigue of carburized differential cross.
Shot peening is known to increase fatigue strength; hence, tests were run to determine the amount of increase to be expected. The above chart shows some of the results. Shot peening was found to provide significant fatigue-strength improvements. Peening surfaces that had suffered grinding damage was found to be very beneficial, 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. The net gain was small. It was also determined that to gain significant improvement the hardness of the shot used was very important. Since carburized surfaces are very hard, the shot must also be hard to be effective.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 93
117
4-35. 4620 Steel: Effect of Nitriding
-- ---- -- -- --
ATMOSPHERE NITRIDED THEN GROUND TO RI1'IOV£ COMPOUND
---
r.~Y~p
Al110SPHERE NITRIDED
QUENCHED AND TEMPERED AT 1050 F (565 CJ
HUHBER OF CYCLES
10~
1') 5
S-N curves for 4620steel, 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, Metals Park OH, 1977, P 286
118
4-36. 4620 Steel: P1M-Forged ksi
100
x
Height strain
80
oE
60
r.-------'-----Time
LI-!
, - - I_ _- ' - -_ _' - - - _ - - - - - ' -
4.0
5.0
6.0 Log cycles
_
7.0
Axial fatigue of P 1M-forged 4620 steel as a function of height strain during forging.
In general, sensitive properties improve as the level of upsetting is increased during the forging process. The diagram above shows the effect for fatigue resistance, although the cyclic stress state also influences fatigue behavior. An interesting feature of P / M-forged parts is the fact that deformation does not significantly affect through-thickness properties as it does detrimentally for wrought material. For re-pressed parts, throughthickness toughness is slightly lower than longitudinal toughness. Upsetting increases longitudinal toughness while toughness in the through-thickness direction remains at a relatively constant level.
Source: B. Lynn Ferguson, "Part II: Fully Dense Parts and Their Applications." in Powder Metallurgy: Applications, Advantages and Limitations, American Society for Metals, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 100
4-37. 4620 Steel: P/M-Forged at Different Levels 700 ,.------,-----....---------,r--------,-------, 100
600 co
a.. :2:
Height strain, HIH o 71%
.'+-----+
x
E
.£ CI) CI)
80
56 & 65%
(JrnaK
500
42%
1'------+ 30%
~
cil
60 400
o~ 0
7.0
8.0
Log cycles
SoN curves for P 1M-forged 4620steel at various levels offorging deformation. As shown, fatigue limit increases as deformation (level of strain) increases.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, p416
119
120
4-38. 4625 Steel: P1M vs Ingot Forms 100
1/
50
Ingot material L,o 192 h 10
1/
/ \
P/M material
L,o 5
o 100
200
V
500
163 1000
h
I
2000
5000
Life, h
Fatigue life characteristics of P 1M roller bearing cups, as shown by a typical Weibull plot. 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.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals. Metals Park OH, 1984, P 620
4-39. 4640 Steel: P1M-Forged 100
,-------.----.r------...-------.-----.------. 0.365-1
':3. 8
80 dir 3-1/4
r-
'"
'Scatter band of SAE 4340
',~,Wrought steel, tested in the '/
"
longitudinal direction
'0
f
" ~
~ u~ c'i)
'0,
,,
9-7/8 R.
60
-,
,, '------------
,,
,
'~
'------_Q:
21
32 Specimen configuration
40 '--------'---------''------'-------'---' 105 10· 10' 103 10' Cycles of stress
R. R. Moore fatigue curve for P/M-forged 4640 steel hardened and tempered to 33 HRC and a yield strength of 138,000 psi.
Water-atomized 4600 steel powder was blended with graphite and compacted in the split punch tooling. Fatigue data for P/M forged 4640 are shown above, and these data fall within the scatter band for 4340 steel. The most impressive statistic is that the P / M-forged parts passed the Army ambient and low-temperature firing endurance tests.
Source: B. Lynn Ferguson, "Part II: Fully Dense Parts and Their Applications,"in Powder Metallurgy: Applications, Advantages and Limitations, American Society for Metals, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, p 103
121
122
4-40. High-Carbon Steel (Eutectoid Carbon): Pearlite vs Spheroidite Property 350. - - - - - - - - , - - - - - - . - - - - - - - , 5 0
45 '" ~. ~
i------j
40
~
'f; c:
35 ~
«
7 6 10 10 Number of cycles to failure
Spheroidite Pearlite
Tensile strength, MPa (ksi) 641(93) 676(98) Yield strength, MPa (ksi) 490(71)(b) 248(36)(c) Elongation in 2 in., % .............. 28.9 17.8 Reduction in area, % 57.7 25.8 Hardness, HB ..... 92 89 (a) Composition 0.78 C, 0.27 Mn, 0.22 Si, 0.016 S and 0.011 P. (b) Lower yield point. (c) 0.1% offset yield strength.
Both pearlitic and spheroidized structures have notably lower fatigue strength than martensitic structures (see 4-1, on p 83). As is shown above, the fatigue properties of spheroidized structures are superior to those of pearlitic structures for eutectoid steels.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 677
4-41. 52100 EF Steel: Surface Fatigue; Effect of Finish and Additives
600
...
4
2
8
6
12
10
500
'"
1 2 3 4 5 6 7 8 9 10
11
.:<
12
.;
..'" 450
....
MLNG MU\ G MLCG SLD G MLNH MU\ H MLCH SLD H MLN P MU\ P MLCP SLD P ~
4.14 3.79 3.45
3.10
til
N
..
~ 400
2. 76
E
~
'" 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
123
124
4-42. 52100 EF Steel: Surface Fatigue; Effect of Surface Finish and Speed
II
600
4.14 3.79 3.45
... 500 Ul
""
. '" rl ..
::: 450
:;
1
2 400
'"e ~
x
350
:2 300
3.10
/---e---i 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
5E 5E 5E 5E 5V 5V 5V 5V 8V 8V 8V 8V 5E 5V 8V 5E 5V 8V
GLl GL2 PLl PL2 GU GL2 PLl PL2 GLl GL2 PLl PL2 GH2 GH2 GH2 PH2 PH2 PH2
N
e
2. 76 ~ to
9
5
2
2.07
10 7 Mean Predicted Cycles to Failure
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
HLN 1.2 HL\ 1.2 HLC L2 ~1.0 1.2 HI.N 112 HL\ 112 HLC 112 ~LO 112
600
.... .:< '"
4.14 J.79
J.45
90% Con fidence Band
'" '" "
~
~
J.IO N
.
::l
400
2. 76
.~ x J50
2.41
"
x
E
~
E
"
.
:E
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
125
126
4-44. 52100 EF Steel: Surface Fatigue; Effect of Lubricant Viscositv, Slip Ratio and Speed
1 2 3 4 5 6 7 8 9 10
600 550 -;: 500
11
.Yo
.
12
MLN MLN MHN MHN SLD SLD MLN MLN MHN MHN SLD SLD
Ll L2 Ll L2 Ll L2 Hi H2 Hi H2 Hi H2
4.14 j.79
) .45
'
:: 450
f---
t
. .
3.10
l/}
N
-!:
:: 400
2.76
G
'" E
~
~ 350
2.41
:>! 300
2.07
250 '-c--'-------'---L-...L.J...J....IJ.Lr--.1_L-LJ...L.I...l..L.L...,-_...L--..L----'-...L-L.LJL..U.,,--........J_-'----.LJ~L.LJ.J 10 5 10 10
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
400
(OMe bl & 68'"
2.76
BASI 011 OA'"
300
2.07
~
vi ~
w
N
~
E
<,
~
\
D<
Co
W
"'4
::c
" " "
z e
200
1.38
.}Is-,f....
:>
~o
o ()"''''
)(
ot
""'..
F
-, -----
'HASiIllOAU. 'I(DI(1I0 BY MUllIPll UQRfUION ANAlYSIS
10
100 l,. LIFE. 10 6 CYClES
Comparison of stress/life data for the mineral oil with and without the ZnDDTP additive in surface fatigue; 85 percent confidence bands for the LSD life estimates are shown and compared with the stress/life relation predicted from regression analysis.
The synthetic and mineral oil no-additive conditions had about the same life. However, the life at all stress levels tested was significantly reduced for the mineral oil with additive below that without additive, by almost a factor of three at the L so level, further indicating a detrimental effect of the ZnDDTP additive on life. Both the synthetic and the mineral oil tests had lives almost two orders of magnitude below the standard Lundberg-Palmgren calculated life. A life reduction factor is used with the Lundberg-Palmgren theory when applied to rolling bearings having high contact angles and thus high slip; but rarely does the slip at bearing contacts approach that level used in these tests, so it is not surprising that the life reductions observed are much greater than the life reduction factors normally used for bearings. The stress/Iife plot shown above is particularly revealing. There is no doubt that the stress / life slope for the additive oil is significantly steeper than for the base stock, which seems to approach the Lundberg-Palmgren theory in stress/fife slope except for the highest stresses where it is even shallower.
Source: w. E. Littmann, B. W. Kelley, W. J. Anderson, R. S. Fein, E. E. Klaus, L. B. Sibley and W. O. Winer, "Chemical Effects or Lubrication in Contact Fatigue-Part III: Load-Life Exponent, Life Scatter, and Overall Analysis," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 285
127
128
4-46. 52100 Steel: Carburized vs Uncarburized 99.
0 ILl
...J
95. 90. 80. 70. 60. 50. 40. 30.
ex 20.
IL.
...
Z ILl
0
10.
a:::
ILl
a.. 5.0 4.0 3.0 2.0 I. 0
L..-_-'------'-----'----'''--'--'l..J...J...L--_-'------'-----'----'''--'--'LJ...LJ
I
10
100
MILLIONS OF STRESS CYCLES
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
CASE---CORE---CARBURIZED----
10'
10·
10'
REVERSALS TO FAILURE,
Summary plots of total strain amplitude: reversals-tofailure data for simulated case, simulated core, and carburized materials.
Carburized material is seen to have low-cycle fatigue resistance intermediate between the simulated case and core material, a common intersection with simulated core material at intermediate lives; and in the long-life regime, carburized material specimens are more fatigue resistant than either simulated core or case material specimens. Plotting the strain-life curves for both case and core simulated materials on a common set of axes , as shown in the above chart, reveals an interesting feature. It has been observed that curves of these materials intersected at a life of approximately 2.NJ = 105 reversals. This is in agreement with the results of this investigation. Intersection of the life curves for simulated case and core materials accounts for a shift of failure location in carburized components.
Source: J. M. Waraniak and D. F. Socie, "Cyclic Deformation and Fatigue Behavior of Carburized Steel," in Wear and Fraclure Prevention, American Society for Metals, Metals Park OH, 1981, P 249
129
130
4-48. 8620H Steel: Effect of Variation in Carburizing Treatments
300
Single --......... reheat .........
•
• -------~rr"~
100
.0
0
o
o
o
L....---L--L....L..JL...U.UJ..._...L......l-JL...L.I..LU.l...----l.---l.....L...L.L.I........ _-'--'-.L..L.J..L.LJ.J---J
103
105
107
Cycles to failure
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
131
4-49. 8620 Steel: Effect of Nitriding
50
§
....
_-- .... _--
~ ~
--
-- .... --
;
.THO~l'\IrRr NITRIDw TNr.N GROUND TO RJ>1OVr COMPO''''" lAnR
AnlOSI'IlERE NITRIDED
:;;
iO QlJ[HCHED AHD TD4PER£D
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
~
10
o o o
8
.....
...
"':'.:.:-. - • _
_
6
. ....
-.-.-.-.
8622 MATERIAL
............... . ...................
4
2
....
..... ............
........
GRINDING WITHam BURl, - • - . - '" SEVERE DAHAGE BY GRINDING •••••••• •••••• (REFER FIG. (~4))
10
10 2 10 3 10 4 10 5 CYCLES TO FAILURE
106
Influence of grinding quality on fatigue properties of carburized differential cross.
C/)
~
z ...... I
§ P-
o 0
0 .....
~
~
t::l
40 30 20
~~
"2'::.".::- •
....-::-.:.......
10 8 6
8822 MATERIAL
.....::: ........
...... ........ ........ ::: . ~
'"
4 2
GRIND AFfER H. T. GRIND BEFORE H.T.
- . _.-
Z ......
ffi ~
10
10 2 10 3 10 4 lOS CYCLES TO FAILURE
10 6
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
4-51. Cast 8630 Steel: Goodman Diagram for Bending Fatigue LEGEND -. - 0 -
1.0
0.9
OoB
X-
- a-
CAViTIES CAST STEEL - SOUND R.R. MOORE FATIGUE SPECIMEN I UNNOTCHED I R. R. MOORE FATIGUE SPECIMEN (NOTCHED 0.001!! in. R I CONTAINED WITHIN THE BAND ARE THE FOLLOWING DISCONTINUiTIES WELD -INCOMPLETE PENETRATION WELD - UNDERCUT WELD - SLAG WELD - MACHINE - SOUND AS WELDED - SOUND SLAG INCLUSION HOT TEARS
0.7
0.6
0.5 l:
l-
l:
I-
e> ~ 0.4
Z III
III
Ul
:n
0:: 0:: I-
;.:.:.:=.::.::.....:.:.:=--=-""-=~
III
III .J
e>
~O.2
=>
~
lL
RANGE
0.3
III
I-
0.1
0
o.s MEAN
- 0.1
STRESS
TENSILE STRENGTH
- 0.2
- 0.3
- 0.4
Goodman diagram for bending fatigue for normalized and tempered cast 8630 steel.
Data here show that severe discontinuities lower the fatigue strength of cast steel. However, it will be observed from the Goodman diagram above that the results of the notched [0.0015 in. (0.0381 mm) radius] R. R. Moore fatigue specimen fall below those of the other bending fatigue values. Goodman diagrams for torsion fatigue and for a quenched and tempered heat treatment show similar conditions with the notched fatigue values below the surface discontinuity values. In many cases, therefore, design, based upon notched R. R. Moore fatigue data, introduces a safety factor. It must be remembered that the discontinuities were very severe and exceeded all ASTM classes of nondestructive inspection standards. The allowable discontinuities described in the ASTM standards are therefore expected to exert a somewhat less damaging effect on fatigue behavior.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed.. Steel Founders' Society of America, Rocky River OH, 1980, P 15-32
133
134
4-52. Cast 8630 Steel: Effect of Shrinkage 0.6
,--------,----,--------,r------, END.
UTS
0.5 (/) (/)
W
!.!l
(~)
138
(951)
138 135 137
(951) (931) (945)
~ 0.17 0.13 0.26 0.28
g:(/)1-::c 0.4 o
WZ
::::>W
00::
i= Ii;
It
0.3
W 0-.J
I:::=0CLASS 2 SHRIN K
6.
z_(/)
•
~ ~ 0.2
""
~
t.
SHRIN~~A~~
Zl-
CLASS 6
0::
( EXTENDS TO ~URFACE )".
w
I-.J <3:
~.
(SUB SURFACE)
flLASS 2 SHRINK 6. (EXTENg~ SURF.)
-
--.-~
0.1
NO FAILURE
0'--------'-----...1.--_ _---' 104 10 5 106 10 7
CYCLES
----'
TO FAILURE
Effect of shrinkage on plate bending fatigue of quenched and tempered cast 8630 Ni-Cr-Mo steel.
As shown in the chart here, 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.
Source: Steel Castings Handbook, 5th Edition. Peter F. Weiser. Ed., Steel Founders' Society of America. Rocky River OH, 1980, P 15-30
4-53. Cast 8630 Steel: Effect of Shrinkage on Torsion Fatigue 0.5
...------,-------r-----r--------, I I I TENSILE STRENGTH
(f) (f)
0.4 -
w
.84 - 91 ksi
-
(579 - 627 MPa)
OCr
t;;t;
oc Z w <1 w OC
0.3 -
°
rt;;
•
(f)
w
~....J Z(f)
CLASS 6
zlOC
W I....J <1
NO FAILURE
....
NO. 15 NO. 13
CLASS 6 SHRINK
x-x
NO. 13
CLASS 2 SHRINK
0-0
0.1 r-
2
~~SHRINK SHRINK-./ -1-.. ......
0.2 I-
~~
rSOUND
x I <. X~ 0 . . . . . . . . . ~- 9t,--:;, O'::CLASS 0 """"--
SOUND
CYCLES
-
TO FAI LURE
Effect of shrinkage on torsion fatigue properties of annealed cast 8630 steel.
Source: Steel Castings Handbook. 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 15-31
135
136
4-54. Cast 8630 Steel: Effect of Shrinkage on Torsion Fatigue 0.5
,....------r----~----...,__---___.
T ENSILE
en en
0.4 I-
W lr
. . XC o ,~?... o Xo ..... _ 0 o 0 o~--o_
Z
W lr II-
«<.
en en
X_X_ o~ 0 00 - - - 0 _
W
~--1
Z
in
(917-951 MPo)
....
o
lr
STRENGTH
133-138 ksi
I-I en I-
I
I
I
0.2 I-
j::Z
-
NO FAILURE
ZI-
0-0
NO. 15
SOUND
lr W
0--<>
NO. 12
CLASS 6
SHRINK
x-x
NO. 12
CLASS
SHRINK
~
0.1 I-
2
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.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, p 15-31
4-55. Cast 8630 Steel: Effect of Shrinkage on Plate Bending
en en W a:: I-J: en t;
~
6,
SHRINK -----
0.4
W
( SUB SURFACE 1
~...J
Z _en
I-Z
a::
STEEL
W
6
I-
...J
AXSOUND
CLASS 2
~a::
Lt
\
0.5
WZ :::>W
~tn
l:i.-,-------,-----,-------.
0.6
0.2
°
L..-
II 14
o_"'\..
°
6 "'0'0 6''00, 0
UTS ~ I MPo 1
83 B4
..J.......
(5721 (5791
----'
'--6_
0_
END. RATIO
.35 .32
.........
.....J
104
CYCLES TO FAILURE Effect of shrinkage on plate bending fatigue of normalized and tempered cast 8630 Ni-Cr-Mo steel.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 15-31
137
138
4-56. Cast 8630 vs Wrought 8640 600,--------.----------,-----------, Notched Unnotched Wrought 8640 0 • 80 Cast 8630 '" '" 5001--------,,------=""""'ct----------t---------; .0;
0: 60 " ::i 400 1-----------+-----2-"""=-+-~---------l ~ ~ ~
:;;
IJ)
300 I-----------+"""-~""'_=------t------------j 40
Normalized and tempered to 220 HB
0.1
1
10
Millions of cycles to failure
600 Ouenched and
80
tempered to
286 HB
500 .0;
~
c,
sc
:;;
e
60 ~.
400
U5
U5
300 40 200L----------'----------'-0.01 0.1
-----J
10
Millions of cycles to failure
The fatigue limit for smooth-machined specimens is generally about one half the tensile strength, but is reduced considerably by notches or a rough cast surface. The S- N curves in the graphs above compare wrought 8640 and cast 8630 steel in two different conditions of heat treatment. In both of these comparisons, the wrought 8640 is superior, but the two steels are practically identical in the notched fatigue test. 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.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, p 397
4-57. 8630 and 8640 Steels: Effect of Notches on Cast and Wrought Specimens Tensile strength. ksi
80
700
100
120
140
160
r---.-,-----..------.-.---.----.----.------y-----,
100
, •
600
Cast steel
I---I-------Ir----t-----t-----..~!i'±i~_+--___l
80
5001---1--
"'----+---t----j
60 ]
"
.~
u.
300
40
20 100
L -_ _-'---_ _-'----_ _--'---_ _--'---_ _-'---_ _-'---_ _---l
500
600
700
800
900
1000
1100
1200
Tensile strength, 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; note curves above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 397
139
140
4-58. Nitralloy 135 Steel: Effect of Nitriding
60
.....
..........
..... ......
.... .....
...... ......
ATHOSPHEAE NITAIDED THEN CADUND TO RDlDVI: CDHPOUND lAYER
---
AntDSPHERE NITRIDED
50
40 QUENCIlED AND mtPERED
AT 1010
r (16' C)
MlHIER OF CYCLES
10
7
S-N curves for aluminum-bearing nitriding steel (Nitralloy 135), gaseous atmosphere 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 287
141
4-59. AMS 6475: Effects of Welding
120 _110 V)
e, ':I. 100 If)
(/) 90 lJJ ~
~
AMS b 4-75 (CE.VM)f ~TI r;.UE R:.R Me OR 1:.- CSMOO Ir-H i)Pec,
"
C
\ "",
----- ;OR~
'\.'~ "
,
WEWDE't>+ST~~~!l
;
<, "--
" '.,
"
80
- - - - ~~ \~EL.\:)ED ( E.B')
.......
I
i
~eL.. i
--+ -- ---- -
i"o .. _ ~--
IO~
106
107
10 8
CYCLES TO FA'LUR~ Fatigue strengths for case-hardened materials as well as through-hardened may be satisfactorily defined using the R. R. Moore rotating specimen test. The smooth unnotched Moore specimen is ideally suited for studying many of the effects of manufacturing and processing variables upon fatigue endurance. An example of the use of this testing technique in the evaluation of electron beam welding and postwelding aging effects upon the endurance limit of basic AMS 6475 material is shown in the above S-N diagram.
Source: Charles W. Bowen, "Review of Gear Testing Methods," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 346
142
4-60. Medium-Carbon, 1Cr-Mo-V Forging: Effect of Cycling Frequency
600 500
'E,'! 0.75 %
400 300
w 500 .-..---r-.-n'TTTT"--'--'-rTT'1~---.-.--rTTTTrr-....,.-;".,.""""" a:: ::> --l400
~
g 300 (f)
w 200 --l U
t>
100
L-..L......L--L.L..I.lllL----'--l-L..L1.JL.U.L...-..L.......L.L.L.U.Lll....---'---'L.J...J.J..WJ
~:] ~': ;,:,~:~ ::~ ::::' : : : :J 10- 2
10- 1
1.0
10.0
100.0
FREQUENCY - CYCLES PER MINUTE
Influence of cycling frequency on the fatigue properties of forged lCr-Mo-V steel at 1049 OF (566°C); no dwell period.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH. 1980, P 15-55
4-61. EM12 Steel: Effect of Temperature on Low-Cycle Fatigue
I
1= 1lef"rmal:ion ""'90-
I-
li'T
(%)
I-
II-
-~
I-
1550°
O/ll---I---+-+-H--+---+-+-H--+---+----I---If-+--+---+-I-H--t---+---I-l--l I-----I-++t--+--+--HH----I--+--t--t-+---t----t-+--HI--+---t--t-t--l
I
10
I
I
r
I
10~
Low-cycle fatigue ofEM12 at 20 and 550°C (68 and 1020 OF).
As holds true for other ferritic steels, the effect of hold time in compression is slightly detrimental to fatigue life.
Source: Philippe Berge, Jean-Roger Donati, Felix Pellicani and Michel Weisz, "Properties of EM 12." in Ferritic Steels for HighTemperature Applications, Ashok K. Khare, Ed., American Society for Metals, 1983, p 114
143
144
4-62. Cast O.5Cr-Mo-V Steel: Effects of Dwell Time in Elevated- Temperature Testing ~
+,'
3.0 , - - - - - r - , . - , - - , - ,r-tr-'r-r-r-t-t-t-r-r-r-r-r-r-r-r-r-......-r-r-r-rr-r-r-r-r-r-r-r-r--r
Z
«
6
~
o "J
-l 1.0
~
REVERSE BENDING MATERIAL A MATERIAL B MATERIAL C PUSH - PULL
°
g
MATERIAL
A
LL
o
............CONTI NUOUS CYCLE
W
t:l
Z
L-_----'-_..L-....l-..I......JL..LLJ...JL-_-'-_-'---...l......l.....J.-L-L..LJ--:-_-'
10 2
10 3
CYCLES TO FAILURE Nf
'f--
___
I
-6-----6_
..... z
~ 0.5 % STRAIN
W 0::
-6----_6
::J
_
-l
~
100
f0-
--- A-=
~ 1.5 % STRAIN
g (f)
W
.J
U
>U I
I
1.0
10
DWELL PERIOD - h Effect of dwell periods on fatigue characteristics of low-alloy cast steel.
As the upper diagram shows, when a D.5Cr-Mo-V steel was tested at 1022 ° F (550 0C), a 20% drop in fatigue life in reverse bending resulted when a D.5-hour dwell was added to each cycle. The lower diagram shows that extended dwell periods, up to 10 hours, have relatively little additional effect beyond that induced by the D.5-hour dwell.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, pp 15-56 - 15-57
4-63. Cast 0.5Cr-Mo-V Steel: Effect of Environment at 550°C (1022 OF) o"e
..,'
"
I
~
REVERSEO BEND TESTS 0. IN AIR + IN STEAM x IN VACUUM
-l 1 0
PUSH - PULL TESTS IN AIR
z
~
°
~
~ u,
o W
o Z «
a::
o
I '-::2--'-----'-...L..JL....L..L.L.L-'--;:---'-----'-...L..J'--'--'c...L.L'--;---'---'--'--'--'-'..........
10
CYCLES TO FAILURE - N
Fatigue endurance behavior of cast 0.5Cr-Mo-V steel at 1022 of (550°C) in air, steam, and vacuum (no dwell period).
Source: Steel Castings Handbook. 5th Edition. Peter F. Weiser. Ed .• Steel Founders' Society of America, Rocky River OR 1980, p 15-55
145
146
4-64. Cast C-O.5Mo Steel: Effect of Temperature and Dwell Period on Cyclic Endurance at Various Strain Amplitudes TEMPERATURE 10 4
100
200
0
c
300
c:---,-----r---,-----.--....--...----,
103 If)
0.6 %
W ...J U
0.7 %
~
U
1.0 %
I
W U Z
1.5 %
=>
0.5 %
0::
0
z
w 10 2
1.0 %
CONTINUOUS CYCLE 30 MIN. DWELL TIME
10 '--_ _. L -_ _- ' -_ _...L-_ _--'-_ _- ' -_ _- ' - - - ' 400 600 800 1000 200 1200
TEMPERATURE - OF Influence of temperature and dwell period on the cyclic endurance of C-O.5Mosteel at various strain amplitudes.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980 pIS-55
147
5-1. HI-FORM 50 Steel vs 1006
~ 0.010
....::::;
::l
o HI-FORM 50
I STRAINED
t.1006
lAND AGED
... ::E
.... ~
VI
U
.....
u >u 0.001 '-3 10
---'10
....L.-:-
4
_
105
REVERSALS TO FAILURE, 2Nf
Strain-life data for AISI 1006 and HI-FORM 50 (a columbium-bearing HSLA steel) in the strained-and-aged condition.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OR, 1981, P 214
148
5-2. HI-FORM 50 Steel vs 1006: Stress Response
0500L
~
~400
~1.FORM50 o-o-~
~~
~ 0.0065 "-"-~C)oO-OC>-_~0.002
gJOO ~
1006
STRAINED AND AGED
OL.-----.L..;-----...I..::------'--;:------L-;------' 2 3 1
10 10 CYCLE NUMBER, N
10~
Stress response of strained-and-aged AISI 1006 and HIFORM 50 steels.
The imposed constant total strain amplitudes are indicated on the graph. The degree of softening of these two steels is less compared with that of dual-phase steels, which simply reflects the significantly lesser degree of strain hardening of the 1006 and HI-FORM 50 compared with the dual-phase steels.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OH, 1981, P 209
5-3. HI-FORM 50 Steel Compared With 1006, DP1 and DP2 500
3=
.......o U
cyo-~
400 0
!
II>
~~ 200
~:;;
«
DP 2
~HI-FORM50
a~ 300 >Uv;
..: ~
_000-0 DP 1
/0
'-1006
100
STRAINED-AND-AG ED
II>
OL...-_...L-_--'-_----'L-_...J..-_--'-_ _
o
.002
.004
.006
.008
.010
CYCLIC STRAIN AMPLITUDE
Comparison offour steels: AISIl006, HI-FORM 50 (a columbium-bearing HSLA steel), a lean phosphorus-bearing dual-phase HSLA steel (DPl), and a carbon-manganese dual-phase HSLA steel (DP2).
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. This confirms the potential for gauge, and consequently weight, reduction that can be realized from the use of higher-strength steels.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OH, 1981, p 212
149
150
5-4. HSLA vs Mild Steel: Torsional Fatigue
300
\,
\\.
A
~ -.
~ I
0: 1'-.~"" " . ~~
-
I spc(o,81) spc(O.81) -'-0-- IAI'I'C4011.111 ..·06-- I APFC45It.Ol ISPcIO.81 --0-- APFC5O(1.01 SPC(O,811
IWIK
A
SPC(1.21)
SPC : MLD STEEL APFC: IIGH STllEHOTH STEEL : THICKNESS(mm) I
~
-.......:;.r......:.~ ~.
...
~ ~.'- • _ D
50
10'
--~
10' NUMBER OF CYCLES
SoN curves showing torsional fatigue of automobile frame steels.
To determine whether the foregoing basic test results apply to the frame models, experiments were conducted. The above chart presents the torsional fatigue behavior of the frame models fabricated with the mild steel (0.8 mm) and each ofthe three high-strength steels. In the high-stress, low-cycle range, fatigue strength differs with the class of high-strength steel but virtually no differences of that nature are seen in the low-stress, highcycle range. The three high-strength steel combinations showed virtually the same torsional fatigue strength values as those of the mild steel (1.2 mm) combination, indicating the possibility of gauge reduction.
Source: M. Takahashi. "Criteria of High Strength Steels for Applying to Automobile Frame Components," in HSLA SteelsTechnology & Applications, American Society for Metals, Melals Park OH, 1984, P 498
5-5. Proprietary HSLA Steel vs ASTM A440 0.03 .-----,----,----T""""--T""""----,
0.01
t---=~~f___t--_t--_+----l
.,
"tl
.e Q.
E '" e
en'i!
Proprietary HSLA 690 MPa (100 ksi] min UTS
0.001
0.0004
t---t------jt---t---t--~
L..-_~L..-_~L..-
102
103
104
_ _L -_ _.L-_
_____I
105
Cycles to failure
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),
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 672
151
152
5-6. Comparison of HSLA Steel Grades BE, JF and KF for Plastic Strain Amplitude vs Reversals to Failure 2.0
•• 1.0
t 'ill-
~N ui
0.1
~
/.Aj-pl= 1.871(2Nf)-o·8396
::J e,
R 2= 0.968
::iE
-e
z·
:(
a: Ien 0.Q1
•
0
i=
en
:5e,
•
• BE(Cb-) • JF(Cb-V) • KF(Cb-V-SI)
0.001 10 2
10 3 10 4 10 5 10 6 REVERSALS TO FAILURE. 2Nf
10 7
Plastic strain amplitude vs reversals to failure for Cb (BE), Cb- V (JF) and Cb-V-Si (KF) steels.
For plastic strain-life relationship the statistical analysis indicates that there are no significant differences between the three steels (F-ratio is not significant). This is further illustrated in the above chart, where all the plastic strain data are plotted as a function of reversals to failure. It is clear that a single straight line can adequately describe all the data. Such a regression line is drawn as the solid line in this chart.
Source: Shrikant P. Bhat, "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steels," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 588
5-7. Comparison of HSLA Steel Grades BE, JF and KF for Total Strain Amplitude vs Reversals to Failure '
.,. w
o
::::II-
2
::::i e, :::l:
<
z
< a:
I-
en
...J
-e
I-
8 12 0.1'------'----'----'----'-----' 10 2
10 3
10 4
10 5
10 6
REVERSALS TO FAILURE. 2N f Total strain amplitude vs reversals to failure for Cb (BE), Cb-V (JF) and Cb-V-Si (KF) steels.
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, the three steels behave similarly and the differences between them are minor.
Source: Shrikant P. Bhat, "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steel," in HSLA SteelsTechnology & Applications, American Society for Metals, Metals Park OH, 1984, P 587
153
154
5-8. Comparison of a Dual-Phase HSLA Steel Grade With HI-FORM 50: Total Strain Amplitude vs Reversals to Failure ~ 1.0
w-
o
...='::::;
0
DUAL PHASE 1
16
AS-RECEIVED STRAINED AND AGED
HI-FORM SO - - -
".
:e
« z
«
...'" '" ...u u
>- 0 . 1 ' - : : - - - - - - - - - - - ' - - : - 4 u 103 10
.1--:10
__
5
REVERSALS TO FAILURE, 2Nf
Total strain amplitude versus life data for DPI (a lean-phosphorus HSLA steel) in the as-received and strained-and-aged conditions. Data for HI-FORM 50 (a columbium-bearing HSLA steel) are included for comparison.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention, American Society for Metals. Metals Park OH. 1981, P 213
5-9. AISI 50 XF Steel: Effects of Cold Deformation 8
'?
...
0
IJo
tJ>
Eeff
I(
A
~
0.2
0.4 00.6
D
.r
s
6
w· 0
00
~
I-
::::i e,
~
4
z
cta:
I-
CIl
...J
2
l-
e I103
104
106
105
REVERSALS TO FAILURE, 2Nf
8 Eeff
'?
... 0
A
I(
~
.r
s
0.20
00.30 6
D
0.57
w·
0
~
!:
...J
e,
~
4
z
ct lii a:
...J
2 0
I-
o I103
104
105
106
REVERSALS TO FAILURE, 2Nf
Total strain amplitude versus reversals to failure for AISI 50 XF HSLA steel. Upper chart: after balanced biaxial stretching; lower chart: after cold rolling.
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, some specific differences were apparent; for example, 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.
Source: John M. Holt and Philippe L. Charpentier, "Effect of Cold Forming on the Strain-Controlled Fatigue Properties of HSLA Steel Sheets," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 218
155
156
5-10. AISI 80 OF Steel: Effects of Cold Deformation
b
8
po
)(
~
-r
s
6
w·
c
::l I-
:::i a-
~
4
•
'lI
Effective Str8in 8nd Mode of Deformation
4 60.0 '90.06 a 0.08 00.16
1 Uniaxi81 T8nsion
881. Biaxial Stretching
Solid Symbols-D8ta for Specimens Tr8nsverse to Hot Rolling Direction
z
« II:
In ..J ~
b
Runoutl
2
I-
REVERSALS TO FAILURE, 2Nf
Strain-life curves after deformation for AISI 80 DF HSLA steel.
In this steel, the fatigue life appeared to remain unchanged or to increase very slightly as a result of deformation, at least for the effective strain levels investigated (see graph). Also, the fatigue life appeared to be unaffected by the mode of deformation and the specimen orientation.
Source: John M. Holtand Philippe L. Charpentier, "Effect of Cold Forming on the Strain-Controlled Fatigue Properties ofHSLA Steel Sheets," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, p 218
5-11. Comparison of Three HSLA Steel Grades, Cb, Cb-Vand Cb-V-Si: Strain Life From Constant Amplitude 2,..,-----------------,
t
t
~
w c :::>
I-
::::i 0. ::l!i
2,-----------------,
~
w
c
0.1
:::>
0.1
I-
::::i 0. ::l!i
4(
~
• TOTAL .PLASTIC AELASTIC
4(
4(
z
a:
:;:
I-
en
a:
I-
en
0.01
0.01
• 10 3 10 4 10 5 10 6 REVERSALS TO FAILURE, 2 N f -
10 7
Strain-life curves for the Cb steel.
0.001 '-----''---------'-----'-----'-----' 10 4 10 5 10 2 10 3 10 6 REVERSALS TO FAILURE, 2 N f -Strain-life curves for the Cb-V steel.
~
0.1
l!f
E ..J
~ 4( Z
:;:
a: 0.01
lii
.1OTAL • PLASTIC AELASTIC
Strain-life curves: Strain-life data from constantamplitude tests for the three steels are plotted in the three charts here respectively in the form of total strain amplitude versus the number of complete reversals to failure.
0.001'----'----'----'----L....:..---' 10 2 10 3 10 4 105 10 6 RI;VERSALS TO FAILURE, 2Nf Strain-life curves for the Cb- V-Si steel.
Source: Shrikant P. Bhat, "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steels," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 583
157
158
5-12. Comparison of Stress Responses: DP1 vs DP2 Dual-Phase HSLA Steels
------ =........... --
600
_11
:
5
0
~
-
1 _
'==... - .... .............
:E ::: ...
:-..
"'- ...:=:.-~ . .....
~_
--;::J"
..- 0.0065 -.:-----0.003
~ .. 009--o_-.>-""""_ _ ~-0.0065
0' 2
~
...~boo ~ v
0---
C
1
:
c
I
°0.0025
~200
STRAINED AND AGED
0'-::10°
'-;-
'-::-
'-::-
---'''--:-
--1
10'
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.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention. American Society for Metals, Metals Park OR, 1981, p 209
5-13. Dual-Phase HSLA Steel Grade: Stress Response for As-Received vs Water-Quenched
..;
0.005
6 u ~ 400'r-o-OCo--o-----..,.--<-->--e..-"d""--ov-...o-O0.002
'" ~ o
~300 v
::::; v
>-
v200
Ol.I
..L..,-
....L..:--
........-::-
--'-:-
--'
10 2 10 3 CYCLE NUMBER, N
Stress response of a water-quenched dual-phase steel in the as-received condition for total strain amplitudes of 0.002 and 0.005.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention. American Society for Metals, Metals Park OH. 1981, P 208
159
160
5-14. Dual-Phase HSLA Steel Grade: Stress Response for As-Received vs Gas-Jet-Cooled
.500 Go
:e
..__
E.. oo
= ~
'"
~
~-----o.--o_o(a o
3 0 0 r - - o - - _....-
I~
....-
0.006
...._ -....._-"''-oA.-o''-_..o-_ _....._0.0025
v
~hoo
. ..>
>v
e
Cl00 C
OL-
----''-;-
----'-:;-
--l.-:;-
--l.-:;-
---'
I
CYCLE NUMBER, N
Stress response of the gas-jet-cooled dual-phase steel in the as-received condition for total strain amplitudes of 0.0025 and 0.006.
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OH. 1981. P 208
162
5-16. Comparison of Dual-Phase HSLA Steel DP2 With HI-FORM 50 0
0.010 DUAL PHASE 2
1'"
AS-RECEIVED STRAINED AND AGED
HI-FORM SO - - - -
0.001L..:---------"---=,.-10 4 10 3
...L..,,---
..J
10 5
REVERSALS TO FAILURE, 2Nf
Strain-life curves for DP2 (a carbon-manganese HSLA steel) in two conditions compared with HI-FORM 50 (a columbium-bearing HSLA steel).
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OH. 1981, p 214
164
5-18. Fatigue Crack Propagation Rate: Effect of Temperature for Two HSLA Steel Grades
HSLA-1 G.S. =lOfLm R =0.1 o 300K t> 233K Cl 173K o 123K Q)
!'/;f/:/.
10 ,
'/:
>.
o "E
I:
g' i
t> 233K
Q)
,
I,
o~
',, (PlI
0123K~I1!I: , I I
<,
b
E
•
01:
o 173K o >. o
1/ /
1//,
o 300K
.
1/ t>A
f
I/,~
R=O.1
/
/?~ I 0,: t>.
u
HSLA- 2 G.S. = 10JLrn
Prof , 6'
01 '
.,pI ,
I , / t>, I
,
Pi
/,/
P,
T(K)
n
300 233 173 123
3.6 5.1 6.5 10.8
T(K) 300 233 173 123
14
n 7.6 8.8 12,1 15.6
18 22 26 30 36
6K, 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.
The only significant difference between HSLA-I and HSLA-2 is that HSLA-2 contains double the amount of Nb that HSLA-I contains (see compositions on p 165). The effect of temperature is seen to decrease the crack propagation rate with decreasing temperature at low values of 6.K. However, as the stress intensity increases, a crossover occurs wherein higher growth rates were observed, as shown in the above charts. This crossover is further reflected in the increase in the Paris law exponent, n, where it ranged from 3.6at room temperature to 10.8at l23K for HSLA-l. The large increase is a result ofthe change in the fracture mechanism from ductile transgranular fracture to cleavage. This behavior has also been seen in iron binary alloys where n increased from 3.5 at room temperature to 20.9 at l23K.
Source: Khlefa A. Esaklul, William W. Gerberich and James P. Lucas, "Near-Threshold Behavior of HSLA Steels." in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, p 569
165
5-19. Effect of R-Ratio and Test Temperature on Crack Propagation of H SLA Steel Grade 1
HSLA-1
HSLA-1
G.S. 10pm T-300K
G.S. 10pm T -123K
• R-O.IO
• R-O.IO R-0.35 • R-0.70
10'
10
z
~
10 R-
0.35 • R-0.70
e.
-,
c lO
"0
4
II
II
7
II 9 10
12 14 It III 20 2
4
24
II
II
7
II 910
1214161920
6K, MPa-m l12
6K, MPa-m"
The effect of R-ratio on fatigue crack propagation behavior of HSLA-l at test temperature of 300 and 123K in the as-received condition.
Compositions of HSLA-l and HSLA-2 Alloy
HSLA-1 HSLA-2
.. .
C
Mn
Nb
SI
0.07 0.06
0.51 0.35
0.014 0.03
0.03 0.03
P
<0.005 0.01
S
Al
0.005 0.01
0.01
Ni
Cr
Fe
0.01
0.01
Rem Rem
Source: Khlefa A. Esaklul, William W. Gerberich and James P. Lucas, "Near-Threshold Behavior of HSLA Steels," in HSLA Steels-Technology & Applications. American Society for Metals, Metals Park OH, 1984, p 571
24
166
5-20. Effect of Test Temperature on Fatigue Crack Propagation Behavior for Two HSLA Steel Grades
HSLA-1
HSLA-2
G.S. IOlJm R-O.I
G.S. 10JLm R' 0.1
e300K .233 K
o
.173 K .123 K
II)
u>u
.....
~
o o
300K 233K 173K 123K
E
4
1I
6
7
.6K,
8 9 10
12 14 16 18 20
MPa-m"2
24
4
5
6
7 8 9 10 12 14 16 1820 24
6K, MPa - m112
Fatigue crack propagation behavior of two HSLA steels tested at temperatures of 300,233, 173 and 123K in the as-received condition.
The only significant difference between HSLA-I and HSLA-2 is that HSLA-2 contains twice as much Nb as HSLA-I (for compositions of the steels, see p 165). Near-threshold crack growth and threshold stress intensities for both steels in the as-received condition are depicted in the above charts for all test temperatures. Comparison of crack growth rates and threshold stress intensities at room temperature indicate that HSLA-2 has a higher resistance to fatigue crack propagation than HSLA-l. The stress intensities amplitude, 11K, for constant growth rates of IO- s and 10-9 ta] cycle are 2.0-2.5 MPa-m 1/2 higher in HSLA-2 than in HSLA-l. The threshold stress intensityl1K'h' is also higher for HSLA-2 (8.0 MPa-m 1/ 2) compared to HSLA-I (5.5 MPa-m 1/2). The 2.0-2.5 MPa-m 1/2 difference in threshold and for the two growth rates clearly demonstrates that there is an inherent difference in the fatigue crack propagation behavior of these two HSLA steels. This difference is also reflected at low temperatures, where HSLA-2 showed lower crack propagation rates and higher threshold stress intensities than HSLA-I. Furthermore, by comparison of threshold stress intensities for these two steels in relation to the effect of decreasing temperature on increasing 11K/I" it was found that the ratios of I1K,h (D t 11K,,, (300K) are the same for both steels.
Source: Khlefa A. Esaklul, WilliarnW. Gerberich and James P. Lucas, "Near-Threshold Behavior of HSLA Steels," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 569
5-21. Stress-Cycle Curves for Weldments of Different HSLA Steel Grades
'II'EI.~E
""'-'~"---dE]r ~ "" ~~ ~ =Jl
.~:~,:,.,
(00
~~~
lotAlERrAl
oOJtrElDl!1
~
I~.~
.
USIW 22->€.....
'..po'- '. -'ed'" , "
IoU5 I
I~
lHlER OYIW.IC lCW> - cotI:XTJ:)N:S
S"mESS - LMTS OF WB.DED SAt.flI.ES
OSIE 3eON
5' 52.
JO~
.=2C~
on
'<:
...... ~ ~--.....--.:. ....., ~ <,
10:
.....,~
....
~ - - - f=:-~ ~-..::.
00
I
I
$~ ,,~
i NJIIB[A
~
CYCLES
.q ~
I--
"
Stress-cycles curves of welded samples of different materials under tension load.
"._.-"---dEl
lHlER OYIW.IC
S"mESS - LMTS OF WB.DED SAt.flI.ES
:;).('..,100
,.-1
"" ,
, " ~.
.
E
i~!
I~
'00 •
LOAD - 00NXT10N9
WATEAW.
I~%
USIW 22J.E"'B
,'·0;;',
.. tiS
T~
...,•.,M"A.E- wac,
'-PO -,
,
OStEJOOn-l
-,
5152.3
...::"" )00
"""
."" F--",,~
",
~~
~
.",
---
"- 1--.0-
.";" c-c
0=
"'''''':
.-
sc.J
I "".,.B£R ~
I
CYCLES
Stress-cycles curves of welded samples of different materials under alternating bending load.
Fatigue data were derived from testing welded samples under tension and bending loads. It was surprising that under both types of load the HSLA steel and the soft unalloyed steel hardly differed in fatigue strength; thus it can be said that the use of HSLA steels is not justified if a component has a weld in the highest-stressed area. An explanation for this is the loss of the thermal-mechanical effect, which is responsible for increased strength, by the heat influence during the welding operation; and it is thought that a higher-strength manganese-alloyed steel, such as St 52.3 (according to DIN), the strength of which results from the chemical composition, would be more favorable in this respect.
Source: Klaus E. Richter, "Cold and Hot-Rolled Microalloyed Steel Sheets in Opel Cars-Experience and Applications,"in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 487
167
168
5-22. Weldments (FCAW): SAE 980 X Steel vs 1006
R=0.1. toO.13" 13.3mml
--6--
SAE-1006. Smooth
-'-0-'- SAE·1006. FCAW
- - . - - SAE-980X. Smooth - - - - SAE-980X. FCAW
- ..----.-........ -..
----.1;,.1---...- __
-----6-
•
-'"l:>"'&--C'IS"" 6-"tl.
_
~
10 5
106
NT' CYCLES TO FAILURE Fatigue properties of smooth and FCA W SAE 1006 and SAE 980 X steels.
The fatigue strengths of the smooth HSLA steel were higher than that of the low-carbon steel. The I06-cyclefatigue limit stress of the smooth SAE 980 X steel was 469 MPa (68 ksi) and that for the SAE 1006 steel was 283 MPa (41 ksi). However, after welding, SAE 1006 and SAE 980 X steels exhibited similar fatigue properties over the 104-1 06-cyclelife range studied. The I06-cycle fatigue limit stresses for FCAW SAE 1006 and SAE 980 X steels were between 114 MPa (16.5 ksi) and 117 MPa (17 ksi).
Source: Kon-Mei W. Ewing. Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue of TIGDressed SAE-980X HSLA Steel." in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 556
5-23. Weldments (TIG): DOMEX 640 XP Steel Welded Joints vs Parent Metal
1000 000 "00 100
r---~
__
600
Parent IIletal
sou 400
TIC-treated bull weld
!
JOO
TlG·treUed fillet weld
~ '.00
Untreated butt weld
Untreated fillet 100
50 '--
.L-
.L-
-'-
105
106
.L..-........
CYClES TO FAILURE
Fatigue strength for DOMEX 640 XP. Standard-Wohlerdiagram (log-log scale) with pulsating load (R=min stress/max stress=O). Sheet thickness 5 mm and ultimate tensile strength 767 MPa. .
For unwelded parent metal the fatigue strength of a steel is improved with increasing static 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. 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. After welding, grinding or TI G-treatment can be used to improve the weld geometry. The notch effect at the weld toe is decreased and the fatigue properties can be improved. Another solution is to place the welds in areas where the stresses are low.
Source: Tony Nilsson, "Formable Hot-Rolled Steel With Increased Strength," in HSLA Steels->Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 259
169
170
5-24. Weldments (FCAW Dressed by TIG): Fatigue Life Estimates Compared With Experimental Data for SAE 980 X Steel
10
103
2
--.::::::::::::
CJ) ~
~
..•••.
-,
(U r =-87KSI .e'!:
en
10
1
TIG DRESSED SAE 980 LAP-SHEAR WELDS Kf = 2.52
I
•
.
J!.
:E ..'=ur = 87 KSI
~
10 2
R = 0.1
•
EXPERIMENT - - PREDICTION
I 3 10
101 105 NT,
106
108
CYCLES
Total fatigue life estimates compared to the experimental data for the FCA WITIG-dressed SAE 980 X steel.
It should be emphasized that life estimates made on the FCAW{TlG-dressed welds were based on geometry changes brought about by TlG-dressing. The other beneficial effects such as removal of slag intrusions and inclusions were not considered. The close agreement between the calculated and observed long-life fatigue properties suggested that the majority of fatigue improvement seen in TIG-dressed joints was attributable to the geometry change. The smaller flank angle contributed significantly to the increased fatigue strengths of TIG-dressed weldments.
Source: Kon-Mei W. Ewing, Pei-Chung Wang, Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue of T1GDressed SAE-980X HSLA Steel," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 563
en
5-25. SAE 980 X Steel Weldment (FCAW): Smooth Specimen vs TIG-Dressed vs As-Welded
SAE-980X R= 0.1, t= 0.13" 13.3mml
--06--- Smooth Specimen
TIG-Dressed _.-(}-.- As-Welded
---0---
~-------4..II
~o o '-'il... 0
__ ~
-----ZP-n---..I\
~00o 0 o
,-.~
0
0
-0....
o Jgt. '1I'o-."'"b-
0
'[jCD-.,_. 00
0 -c._
o
'-
''''"00'_
10 5
,-"",..It.0
0
o'-'n-
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, From these data, a significant improvement in fatigue characteristics can be obtained by TlG-dressing the welds,
Source: Kon-Mei W. Ewing. Pei-Chung Wang, Frederick V. Lawrence. Jr., and Albert F. Houchens, "Weld Fatigue of TIGDressed SAE-980X HSLA Steel," in HSLA Steels-Technology & Applications, American Society for Metals. Metals Park OH, 1984. P 558
171
172
5-26. SAE 980 X Steel Weldment (FCAW): Lap-Shear Joints
en
<]
10 1 SAE980 LAP-SHEAR WELDS Kfmax =3.49 I R = 0.1 • EXPERIMENT PREDICTION
105
106
NT, CYCLES Total fatigue life predictions and experimental results for the FCAW, SAE 980 X lap-shear joints.
Source: Kon-Mei W. Ewing, Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue of TIGDressed SAE-980X HSLA Steel,"in HSLA Steels-Technology and Applications, American Society for Metals, Metals Park OH. 1984, p 562
5-27. Microalloyed HSLA Steels: Properties of Fusion Welds
STRESS - LMTS . v; ...~
:.~"",.TC'SS'O\·T~ST
.~
·.• ':._':'0
a: WELDED SN.f'lES
"-'-dEl 10
)10.5
.
~~ ::.
"
~
'0(>0'·
-= lC~
nn
'l;: I~:
'~
oi~hO£l
~~
-;.'
I
051E 38CN
, , ",
>l
"3
:.-........-.:..: ~ ......,
'-.I,!~
~ ..... -...;:..-" ~~ .............
I
I
l:SIW n."E."~'"
-.
I
00
--
'---
,
,
......l[R'.l1.
",-
"00',
101"
'00
iec
COIClIJ1DN8
Lp"o -
I~""
I
I~
lHlER OYNAlolC
w(lO~O'""
q
c«
II
I
lA.IM8ER(JFC'fC~t'S_
I
Stress-cycles curves of fusion welded samples of different materials under tension load.
STRESS - LMTS
..
".~' C' lQl.Orh(i· BfK>'Mj·Y«=ST
1.C.... '00
.. n
~~: ,co
lHlER OYNAlolC
a: WELDED SN.f'lES
"-"-dEi 10
Lp"o -
COIClIJ1DN8
~~~-"£'S~~n
)IQ.S
I~%
'. o~·
,.11'
' 10&1 0'"
'-~
IUTERlAl
~%
USIW 22."'£.',
-.
OSIE380T'4
",
SI S23
'-="
'00
""""
'" f"........C'l>~
'00
~~
~
I
-
".,
,
I
I
--~
~--'=
=---=
-
r-''!!::='- ..
"=
--
-II
III I
fUo'8(A OF CVCl£S-
~
Stress-cycles curves of fusion welded samples of different materials under alternating bending load.
Due to the preferred crack location in the welded areas, it seemed necessary to examine the dynamic strength of fusion welded joints of HSLA steels in more detail, testing the steel used for the crossmember at a minimum yield strength of 380 N /mm 2, in comparison to a soft unalloyed, hot rolled steel sheet. Tensile load and alternating bending load were the selected types for dynamic test. The above charts show the respective stress-cycle curves.
Source: Klaus E. Richter, "Cold and Hot-Rolled Microalloyed Steel Sheets in Opel Cars-Experience and Applications,"in HSLA Steels-s-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 487
173
174
5-28. Microalloyed HSLA Steels: Properties of Spot Welds
LOAD AMPLITUDE 6P, kips AND NUGGET ROTATION
68 N , degree
5.0 0.2 1.0 0.5
GroupC
Fatigue test results for the 2.06 mm (0.081 inch) thick sheet with various weld diameters.
LOAD AMPLITUDE 6P. kips AND NUGGET ROTATION
68 N • degree
5.0 2.0 1.0 0.5
LOAD AMPLITUDE 6P, kN
5.0 2.0 1.0 0.5
Group K
4
5
6
7
LOG CYCLES TO FAILURE, N,
Fatigue test results for 1.02 mm (0.040 inch) thick sheet of different stiffnesses.
Results of spot-weld fatigue tests are presented in the four plots (above and on the facing page) for the stated conditions. Each curve shows the load amplitude, I1P, and nugget rotation values, 116 N' for each test as a function of cyclic life. Straight lines were fitted through the data.
1·~oe""",roOlI"
LOAD AMPLITUDE dP, kips AND NUGGET ROTATION d(~N' degree
o
5.0
175
I
~::S';=I~~:"
LOAD AMPLITUDE dP, kN
2.0
1.0 0.5
Group E
5.0
2.0
1.0 0.5
5.0
0.2 1.0 0.5
d0 dP
Fatigue test results for variations in specimen width and thickness.
5.0 LOAD AMPLITUDE dP, kips AND NUGGET ROTATION :>SN. degree
2.0
1.0 0.5
~ -- -
o
5.0
2.0
1.0 0.5
-
"W-IOI61M11~O"1 ' 02 _ " " . ' 0-66In"lf02fj1fl1
::~' GroupN
~
5.0
2.0
1.0 0.5
de dP
Group 0
5.0
2.0
1.0 0.5 1,.1 0211Yn1004..., , W .. l01611'WT1140n,
0 .. 66 ........,0416 ... ,
dP
de
4
5
6
7
LOG CYCLES TO FAILURE. Nf
Fatigue test results for 1.02 mm (0.040 inch) thick sheet with single and multiple welds.
Source: James A. Davidson, "Design-Related Methodology to Determine the Fatigue Life and Related Failure Mode of SpotWelded Sheet Steels," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, p 542
176
6-1. HY-130 Steel: Effect of Notch Radii
~::;; 800
w·
g 400
I-
:J
Q.
9.5mm
~ 200
6.4 3.2
~ tii
100 80 ~ 60
0:
«z
~z
1.6
0.80
40
0.40 0.20
20 2
3 4 5 6 8 10
2
2
3 4 5 6 8 10
2
3
3 4 5 6 8 10
CYCLES TOFATIGUE-CRACK-INITIATION, NjXI03
Cycles to fatigue-crack initiation versus nominal stress amplitude, for notched specimens with various radii of curvature.
~a,
~:::!: 4000 3000 2000 1000
~ 800
"
~
p=0.20mm
600 500 400 300 200 2
3 4 5 6 8 10
2
3 4 56 8 I
2
3 4 56 8 I~
CYCLES TO FATIGUE-CRACK-INITIATION, NI xI0 3
Same data as in upper graph but plotted versus Ill(
/.JP rather than ~a.
Curvature ofthe notch and I:1Kis the stress-intensity amplitude computed for an imaginary crack whose length is the same as the notch depth, a. Barsom and McNicol used this parameter to compare N j , the cycles to fatigue-crack initiation, in HY-130 steel for notches of constant depth but various radii of curvature. The results are shown in the above graphs. In the upper graph, N, is plotted versus l:1a, where N, is defined as the number of cycles to give a 50-J..!m side notch. There is a wide spread in the curves. As expected, the sharpest notch, lowest p, gave the most rapid initiation at a given stress. The lower graph shows I:1K/ vp plotted versus N;. A narrow-spread family of curves results; these converge as the value of I:1K/vp is decreased to a threshold value I:1K/ vp I'h' the minimum value to initiate fatigue cracks in notches.
Source: M. E. Fine and R. O. Ritchie, "Fatigue-Crack Initiation and Near-Threshold Crack Growth," in Fatigue and Microstructure, American Society for Metals, Metals Park OR, 1979, pp 256-257
177
6-2. 300 M Steel: Effect of Notch Severity on Constant-Lifetime Behavior Minimum stress,ksi
300
s:::;:
~
1500
~.
~E
200 1;; E :> E
:>
.~
E
.~
::;:
::;: 1000
100 Notch
500
severity (107 cycle IifeUmel
_~00'::0:----...,."..::-----,.L..._-_--L._------"L-_-_.L-_-_--L._-----,,------..,..L...------'25~0 -1500 -1000 --500 2000 1000 1500
Constant-lifetime fatigue diagram for 300 M alloy steel, hardened and tempered to a tensile strength of 1930 MPa (280 ksi). Solid lines represent lifetimes obtained from unnotched specimens. Dashed lines represent lifetime of ten million cycles for specimens having the indicated notch severity.
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals. Metals Park OH, 1978, P 670
178
6-3. TRIP Steels Compared With Other High-Strength Grades
IlXXl
1200
1400
MN/m2 1600
1800
2lXXl
2200
1400
200 R • 0.1 TRIP ~175
1200
i
~
u ~150
.,.
SR 4340
IlXXl"/;
i
~
~125 v;
.
4l«l
800
:J
.2'
.l!loo 600
Fatigue strength at 10' cycles (R = 0.1) vs ultimate tensile strength for TRIP steels compared with other high-strength steels.
Studies on fatigue-crack propagation (FCP) conducted under controlled stressintensity amplitude (~K) conditions indicate that deformation-induced transformation retards crack growth in lower-strength metastable austenites, particularly at low ~K, and also exerts a beneficial influence in high-strength TRIP steels, although to a much lesser extent. This growth retardation may be due to crack-closure effects arising from the transformation volume change, which may be particularly effective in the fatigue-threshold regime. Smooth-bar fatigue properties appear to be dominated by transformation hardening, which is desirable under stress-control conditions (reducing strain amplitude) but generally undesirable under strain-control conditions (increasing stress amplitude). In lower-strength austenites, transformation reduces fatigue life under conditions of controlled plastic strain amplitude; under controlled total strain amplitude, transformation is detrimental to low-cycle fatigue life, but a small amount of transformation may be beneficial at high cycles. Similarly, the lowcycle fatigue properties of high-strength TRIP steels are found to be degraded by transformation under controlled total strain amplitude. Under stress control, the fatigue life of lower-strength austenites is greatly enhanced by transformation; for a stress ratio (R= amin/ a m• x ) of 0, fatigue limits in excess of the yield strength are observed. Investigation of the smooth-bar fatigue properties of high-strength TRIP steels at R= 0.1, in which thermodynamic stability was varied by heat treatment, also revealed transfermation enhancement of fatigue life. Such enhancement allows the achievement of exceptional fatigue strength at high ultimate strength levels, as illustrated by comparison with other high-strength steels in the above graph.
Source: G. B. Olson, "Transformation Plasticity and the Stability of Plastic Flow," in Deformation, Processing, and Structure, George Krauss, Ed.. American Society for Metals, Metals Park OH, 1984, p 419
6-4. Corrosion Fatigue: Special High-Strength Sucker-Rod Material 10.000
.
Q,
UI UI
80,000
50,000
.... a:
lUI
40.000
B
....0
::i
lL lL
ce
F
30.000
AB- IN AIR C0 - IN ACID BRINE EF- IN ACID BRINE WITH INHIBITOR
20,000
0
10.000 102
10~ t04 NUMBER OF CYCLES
10' FAILURE
10'
to'
Effect of corrosion and corrosion inhibitors on the SoN curve for highstrength steel (sucker-rod material).
After the first brittle crack is initiated, No.2 is the slow step in the process and electrochemical action is the slowest part of this step. Thus, the effect of corrosion can be illustrated with curves of stress vs logarithm-of-number-of-reversals-to-failure for sucker-rod steel. Corrosion accelerates cracks propagation, so the fatigue curve drops from AB to CD, as shown in the graph. Deceleration of the slow stage with a corrosion inhibitor will raise the S-Nfatigue curve from CD to EF.
Source: Joseph F. Chittum, "Corrosion Fatigue Cracking of Oil Well Sucker Rods," in Corrosion: Source Book, Seymour K. Coburn, Ed., American Society for Metals, Metals Park, OH, 1984, P 380
179
180
6-5. Corrosion Fatigue Cracking of Sucker-Rod Material
.8 Ul
~.6
0
~
Ii 0
ct:
~.4
... 0 :J:
I-
I I
~.2
I
UJ
II
oJ
I I
II
0 FIRST STAGE
SECOND STAGE
!THIRD,
ISTAGEI
0 RELATIVE NUMBER OF CYCLES, PERCENT OF FAILURE
Corrosion fatigue cracking of sucker rods.
This graph shows typical progress of a crack at high stress plotted against number of cycles, showing stages in the fatigue process. Observations of sucker-rod crack penetration as a function of reversal accumulation are possible using a bending apparatus and a magnetic fluorescent powder technique. Penetration vs reversal curves resemble the one shown above when the stress is well in excess ofthe endurance limit. During bending, no penetration is apparent in the first 40-60% of the specimen's fatigue life, even though intrusions and extrusions may form earlier. A crack eventually appears and progresses through the specimen. When the penetration reaches a certain percentage of the cross section, the cracking accelerates until catastrophic failure occurs.
Source: Joseph F. Chittum, "Corrosion Fatigue Cracking of Oil Well Sucker Rods," in Corrosion: Source Book, Seymour K. Coburn, Ed., American Society of Metals, Metals Park OH, 1984, P 378
181
6-6. Hydrogenated Steel: Effect of Baking Time on Hydrogen Concentration
300
Normal Notch Strength = 300,000 psi
o.. ~
275
i---::::::::::::
o~
250
200
(f) (f) Q)
...
175
( /)
"0
150
Q)
a. a.
i'...
-, ~
\
125
\
100
..1
-,.
75
\ <,
1
+
Bake ---
24 h r -
Bake 18 hr
Bake 12 hr
Bake 7 hr Bake 3 hr
~
•
Uncharged +-+-
,~
1
\
:---:-
0\ \
\ -\ ~
l
+ +
"
.-
\
-\
~.
~~
"\\'\
a. 225 (f)
0 0 0
i'-..
Bake 0.5 hr
--
<-
1-
50 0.01
0.1
1 10 Fracture Time, Hours
100
1000
Static fatigue curves for various hydrogen concentrations obtained by baking different times at 150°C (300 OF). Sharp-notch specimens. 230,000 psi strength level.
These are, in essence, static fatigue curves, and the lower critical stress may be considered a static endurance limit-that is, a stress below which failure will not occur for an indefinite period of time. This behavior is sensitive to hydrogen concentration as shown above, where it may be seen that all delayed-failure parameters-notch strength, rupture time, and static fatigue limit increase with decreasing hydrogen concentration. Also, even after 24 hours at 150°C (300 OF), there is still a substantial stress range, of the order of 60,000 psi, over which delayed failure will occur. In an unnotched specimen, full recovery ofthe ductility as measured by the reduction of area can be attained in less than 20 hours at 150°C (300 OF), yet delayed failure will occur after 24 hours or longer of baking time at 150 °C (300 OF).
Source: Alexander R. Troiano. "The Role of Hydrogen and Other Interstitials in the Mechanical BehaviorofMetals,"in Hydrogen Damage Source Book, Cedric D. Beachem, Ed., American Society for Metals, Metals Park OH, 1977, P 154
182
6-7. Hydrogenated Steel: Effect of Notch Sharpness 300 275 250 en 0..
0 0 0
225
T~. I~~ 1\ '\
200
en en
Q)
L..
175
( J)
"0 Q)
0.. 0..
«
150
-a_ ""'"
0 01'\.
a
~\
\
125
~\\
Notch
Radiu~ = 0.010 in.
It-- - - -
>--
.
~.
75
; a1lus J. = 2'In.
>--
>--
Notch Radiu's = 0.020 in.
-\..
\
\
100
Notch
\\
i'\
Notch Radius = 0.25 in.
Notch Radius = 0.001 in .
•
.....-
50 0.01
0.1
I
10
100
1000
Fracture Time, Hours Static fatigue curves for specimens of different notch sharpness. Baked 0.5 hour at 150°C (300 oF).
The variation of lower critical stress with notch severity is shown in this diagram. It is evident that the static fatigue limit rises as notch severity (radius) decreases for hydrogen-charged high-strength steels (using the same baking time).
Source: Alexander R. Troiano, "The Role of Hydrogen and Other Interstitials in the Mechanical Behavior of Metals," in Hydrogen Damage Source Book, Cedric D. Beachem, Ed., American Society for Metals, Metals Park OH, 1977, P 155
7-1. O.5%Mo Steel: Effect of Hold Time in Air and Vacuum at Different Temperatures + 10 I
-
CYCliNG
KEY
CONTINUOUS
..J
< e
l-
....
....... ' ........
lLL
o
loO
w
Z
l.' < Z IX: IX: I-
<
VI
r
w
VI
~
10 0
.. =:::::.:.~.........:.~ . ... :--;: '
30 on HOLD
"
.',~""--............:..
AIR VAC ...
~
..........,,:~:-:-: .... " .""'" ".. . . ..«:»r. . "=':..,:-:. . :--;; -
I::NV. TEMP. AIR 275 K VAC . .. AIR 775 K VAC. II
--.
__
.....::--........
._~.
CYCLES TO FAILURE Effect of hold time in air and vacuum upon the fatigue endurances of a O.5%Mo steel at 275 and 775 K.
Source: R. H. Cook and R. P. Skelton, "Environment-Dependence of the Mechanical Properties of Metal sal High Temperature," in Source Book on Materials for Elevated- Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 83
183
184
7.2 DIN 14 Steel (1.5 Cr, 0.90 Mo, 0.25 V): Effect of Liquid Nitriding 1000
'\ 900
- 140 :'\.
""-8 800
700
I'
.
~ 600
.s
13
IS
- 120 - 100
0A
- 80
';;; -'"
.
~
.~
13
>-
U
500
400
300
- 60
\
\ ~
\
105
C
106
- 40
107
Number of load cycles
Effect of nit riding on fatigue behavior of DIN 14 CrMoV 69 steel (0.14 C, 1.5 Cr, 0.90 Mo, 0.25 V). Curves A and C are for hardened and tempered (not nitrided) specimens; Band D are for liquid nitrided specimens. A and B are for smooth specimens; C and D are for notched specimens
K,=2. Nitriding introduces residual compressive stresses at the surface of steel parts; these residual stresses, together with the increased strength of the nitrided layer, increase the fatigue resistance of the part. The increase in fatigue strength that results from nitriding is illustrated in these S-N curves.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 541
7-3. 2.25Cr-1.0Mo Steel: Influence of Cyclic Strain Range on Endurance Limit in Various Environments 10.0 8.0 6.0
KEY •
U
X
3.0 2.0
z ..:
a: 1.0 fVI
u
0.8
-.
~XX
•
ENVIRONMENT Na (30 ppm OF °21 AIR HELIUM
•
Na \300 ppm OF 02)
. ,
x~
'~
'U ...~~~.
........~
-~
.J
u 0.6 >u
....
11K
•
- .. ~
.........=-
0.4 0.3 0.2
C.I 10~
CYCLES TO FAILURE
Influence of cyclicstrain range upon fatigue endurance of 2.25Cr-l.OMo steel in sodium, air, and helium at 865 K. (Cycle used was approximately up 5 s, hold 5 s, down 5 s, hold 5 s.)
Source: R. H. Cook and R. P. Skelton, "Environment-Dependence of the Mechanical Properties of Metals at High Temperature," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, P 83
185
186
7-4. 2.25Cr-1.0Mo Steel: Effect of Elevated Temperature
Testing temperature
~C o 425 , ?f!. OJ' "Q
• 540 595
2
l1
~
°FI
800, 1000 1100
.'=
a. E c 'iii
'"
J5 0.2 0.1 2 10
103 Cycles to failure
The results of strain-controlled fatigue tests of 2.25Cr-l.OMo steel at 425, 540 and 595°C (800, 1000 and 1100 OF) on specimens of annealed 2.25Cr-I.OMo steel are presented in these S-N curves. Within this range, test temperature had relatively little effect on number of cycles to failure.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 659
7-5. 2.25Cr-1.0Mo Steel: Effect of Elevated Temperature and Strain Rate Temperature,Oc
e
~
104
:e B :G > u
u
103
500
450
400
350
550
Temperature,OC I
600
700
I
I
800
900
1000
1100
Temperature, of
Effect of elevated temperature on strain-controlled fatigue behavior of annealed 2.25Cr-l.OMo steel.
Strain-controlled fatigue tests have also shown (note above) that reducing carbon content to 0.03% results in a reduction in fatigue strength. Furthermore, because of variations in strain aging effect, specimens from one heat with a higher carbon content ran longer at 427°C (800 OF) than at 316 °C (600 OF).
Source: Metals Handbook, 91h Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, p 659
187
188
7-6. 2.25Cr-1.0Mo Steel: Effect of Temperature on Fatigue Crack Growth Rate 6K, stress intensity factor, ksi..Jiil.
10
20
60
I--+-------+-------+--------+-----i
10.4
Testing temperature
10-31--+--oC• 205 o 370
of - - - + - - - - - - - + - - - - - - - - - - - f + - - - l 400 700
1---+-6-455-850---+--------+--------,.~____J<-----,_F____I1____l
6595
1100 10-6
Q>
U
Q>
~
U
~ .£
E 10-4 E
:i
~
\!
s: ~
...e
s:
6
~ 0
l;,
.>< u
.><
eu
10.6
:e'
l;l
1;
:e'
..
.
:!'!
:!'!
"D
"D
10-6
10
20
60
s«, stress intensity factor, MPa vmm Variations in fatigue crack growth rate with test temperature for specimens of 2.25Cr-l.OMo steel tested in air.
Specimens were subjected to cyclic loading at a constant maximum load. Stress ratio was 0.05; cyclic frequency was 400 per minute. As shown, the stress-intensity factor range increased as the crack length was increased.
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978. P 660
7-7. 2.25Cr-1.0Mo Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate b.K, siren intemity laClor. ksi...rm:
2
5
2 I 0-4
4C"7f'/
5
5
/,r:/
/ ! vr:~400C~ /
5
I
~
40 CfKO
2 Frequency, cvcles/mln o 4
1
• 40
6
2 2
5
ilK. stressintensity laflOr, MPaym
(.1 IJJ.K. stress intensity factor, ksi 2
5
vrn: 2 I
2
1-//
3
.
~E
5
E
2
4C~
II; t
III
1/
5
2 I
5
40c"m
400c:pm
2 Frequency. cycles/min o
4 • 40
I
5
2 2 5 4K. slreu intensily 'actor. MPa V;;
Ibl
Data shown above indicate that in elevated temperature tests at a given stress-intensity factor range, crack growth rate increases as cyclic frequency is decreased. These fracture mechanics data may be applied to the design of structural components that may contain undetected discontinuities, or that may develop cracks in service. Stress ratio was 0.05. (a) Tested at 510 °C (950 OF); (b) tested at 595°C (1100 OF). Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society ForMetals, Metals Park OH, 1978, p 661
189
190
7-8. 2.25Cr-1.0Mo Steel: Fatigue Crack Growth Rates in Air and Hydrogen 10-5
I
~
10-6
v i>. i>.V
vi>. Vi>. i>.
~ -.
E
z·
.Jl i>.
10-7
-.
'"
al
i0
.dt
10-8
.
°•
Ol
o~ , o ,
U
Gl :::I
-
•
oOe
.l'
~ u
-
0' oil'
't:l
.....
~
~
~
't:l
'" s:
;.
-
••
D
Gl
1i
,•
10-9
0°.,.
l-
Ol .;::;
'"
u..
8 10- 10
I
3
4
Frequency
•
•
§
f-
10- 11
f:
2% Cr - 1 Mo Steel R = 0.05
• • •
V i>.
Environment Air
2 Hz, 50 Hz
H, °0 50 5 Hz
-
}
-
138 kPa H2
2 Hz 0.5 Hz
~I Threshold
,f
I
5
6 78910
I
I
J
I
20
I
J
40
I
j
60
I
I
I
80 100
Stress-intensity factor range, L1K, MPa . m 1/2 Fatigue crack growth rates in 2.25Cr-l.OMo steel in air and in hydrogen.
Corrosion fatigue descriptions are further complicated by the fact that the environment may produce multiple effects. For example, Suresh et a/demonstrated that dry hydrogen may produce afrequency-sensitiveenvironmental effect analogous to SCC at intermediate t:J.K values and a frequency-insensitive environmental effect near the threshold. This is illustrated in the above graph for 2.2SCr-1Mo steel tested in air and in 138-kPa hydrogen gas. Because the sustained-load threshold for this steel is on the order of 90 M Pa . m 112 (82 ksi . in. 112), the K""'j of about 22 MPa . . 1/2) gives • K th : v. m 1/2 ( 20 k' SI . Ill. K ,h ({) ...? ~ lt can be seen for t:J.Kvalues greater than K ,h ({) that there is a large increase in growth rate for the low test frequencies but not for the higher ones. Therefore, this regime may be considered to be one where superposition might apply. In addition, however, there is a true threshold, t:J.K,h , which appears to be frequency-insensitive but which nevertheless decreased by about 30% to SA MPa . m 112 (4.9 ksi in. 112 ) because ofthe hydrogen environment. Such mulitple effects are poorly understood and are clearly possible in a large number of material/environment systems.
Source: W. W. Gerberich and A. W. Gunderson, "Design, Materials Selection and Failure Analysis," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for Metals, Metals Park OH, 1982, P 333
191
7-9. 2.25Cr-1.0Mo Steel: Effect of Holding Time
120days
c
E
Compressive or tensile hold Type of strain hold
Strain range.%
Compressive
Tensile
Both
2.0
loJ
104 Cycles to failure
1.0
C.}
0.5
• •
:::2 } .
o
X indicates zero hold lime
Cycles to failure
(b)
Time-to-failure/cycles-to-failure diagrams for annealed 2.25Cr-l.0Mo steel tested in strain-controlled cyclic loading at (a) 480°C (900 OF) and (b) 540 °C (1000 OF). 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). Strain amplitude indicated by shape ofsymbols and figures along zero-hold-time line.
In these "time-to-failure/cycles-to-failure" diagrams, the lowest curve (zero hold time) indicates the corresponding time period and number of cycles to failure for continuous strain-controlled fatigue tests over the strain range from 0.4 to 2.0% with no holding period at maximum strain. The other curves, which are approximately parallel, are for increasing periods of holding time at maximum strain levels in either tension or compression. The vertical curves are drawn through the number of cycles to failure for each particular cyclic strain. For all tests at 2% strain, failure occurred in less than I000 cycles regardless of holding time or whether the stress was tensile or compressive. The effect of reducing the strain increment and increasing the holding time on number of cycles to failure can be determined from the appropriate curves in the figures.
Source: Metals Handbook, 9th Edition, Volume I. Properties and Selection: Irons and Steels, American Society for Metals. Metals Park OH. 1978. pp 662-663
192
7-10. Cast 2.25Cr-1.0Mo Steel, Centrifugally Cast: Fatigue Properties at 540°C (1000 OF) 60,------r----r---r---r--r---.,.------,
\
5.0 • • •
I 4.0 lLJ
o Z
cr
•••
3.0
z
cr
I-
(f)
2.0
...J
~
g 1.0
OL...._ _-L-_ _---1._---1._.l.-...L-_ _--'-_ _- . J
100
200
400 600 1000 2000 CYCLES TO FAILURE
4000
Fatigue properties of 2.25Cr-1.0Mo centrifugally cast pipe, A2l7, Grade WC9, at 540°C (1000 OF).
Source: Steel Castings Handbook, 5th Edition. Peter F. Weiser, Ed.• Steel Founders' Society of America. Rocky River OH. 1980, P 15-55
7-11. H11 Steel: Crack Growth Rate in Water and in Water Vapor
0.5 ,........----,-----,--"'"""'T--"'"""'T--~--....,
"0
0.3
c E 0.1 ..... c
Q)
.05
H -II Steel
230 ksi Y. S. K - 30 ksi IT'"
o Activation Energy 9,000 cal./gm-atom
o a:::
-...
s: .03
•
~
o
o
.Jl: U
...
o u .01
o Water o Relative Humidity 100% at Test Temperature •
0.005
II
II
II
II
80° F,
Tested at Higher Temperature
Crack growth rate versus temperature for an HII steel in water and water vapor.
It is of considerable interest that the strain rate and temperature dependence of hydrogen embrittlement, as determined by ductility measurements after rising load tests on hydrogen-charged materials, show a characteristic behavior that resembles closely that seen with crack growth rate measurements and external hydrogen environments.
Source: Herbert H. Johnson. "Keynote Lecture: Overview on Hydrogen Degradation Phenomena," in Hydrogen Embrittlement and Stress Corrosion Cracking, R. Gibala and R. F. Hehemann, Eds., American Society for Metals. Metals Park OH. 1984, P 18
193
194
7-12. 9.0Cr-1.0Mo Steel: Creep-Fatigue Characteristics
9% Cr 1% Mo AT 525°C WITH OR WITHOUT TENSILE HOLD TIMES TO hOWELL (TOTAL STRAIN RANGE; FROM (22))
t
LOW C 9% Cr 2% Mo AT 550°C CONTINUOUS CYCLING (TOTAL STRAIN RANGE; FROM (2311
~ UJ l-' Z
« 10°
cr
z
~
cr ~
Vl
9% Cr 2% Mo NbV AT 550°C (PLASTIC STRAIN RANGE; FROM (24))
~--5
MIN. TENSILE
DWELL
CYCLING
10' CYCLES
TO FAILURE
Illustrating the elevated temperature low-cycle fatigue and creep-fatigue properties of normalized and tempered 9% Cr Mo variants.
In this chart are presented the elevated-temperature-fatigue and creepfatigue data for the 9%Cr-1%Mo steel as a single curve in terms of total strain range against cycles to failure; also shown for direct comparison are the continuous cycling fatigue data for the low-C, 9%Cr-2%Mo variant which, although inferior at relatively high strain ranges, suggests superior endurance may be attained in the high-cycle region. From the limited evidence, it seems probable that normalized and tempered 9%Cr-I%Mo steel may be used in reactor-quality sodium at service temperatures with little effect on tensile properties and stress rupture strengths or ductility and that the short term low-cycle fatigue endurance will be increased and fatigue crack growth rate reduced. This behavior is a consequence of the structural stability of the material with respect to interstitial element transfer in liquid sodium and also the low oxygen potential of the overall system which may be expected to preclude oxide penetration and enable partial recohesion of the crack faces during fatigue.
Source: S. J. Sanderson, "Mechanical Properties and Metallurgy of 9%Cr I%Mo Steel," in Ferritic Steels for High-Temperature Applications, Ashok K. Khare, Ed., American Society for Metals, Metals Park OH, 1983, P 95
7-13. 9.0Cr-1.0Mo Modified Steel: Stress Amplitudes Developed in Cycling
-------------------------200 t - - - - - -__
en w a::
------ --- .. ----Continuous Cycle
100 e a.. ::E
Fe 9Cr IMo BV, Nb
o
(/)
l-
(/)
-100 - - with 30 sec T H
---
-200
r,.,..,,"'="::=-:-:.....---......:_:'":_;:"::-=-=_,,:_::- - - _... 10
100
1000
10000
CYCLES This chart shows stress amplitudes (tensile and compressive) that developed in the course of cycling the modified Fe-9.0Crl.OMo steel through a total strain range of 0.5% at 649°C (1200 OF). Fatiguing was carried out in vacuum. Dotted curve indicates continuous cycling; solid curve indicates cycling with a 30-s hold at maximum tensile strain.
Source: S. Kim, J. R. Weertman, S. Spooner, C. J. Glinka, v. Sikka and W. B. Jones, "Microstructural Evaluation ofa Ferritic Stainless Steel by Small Angle Neutron Scattering," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf, Eds., American Society for Metals, Metals Park OH, 1984, p 175
195
196
7-14. 9.0Cr-1.0Mo Modified Steel: Effect of Deformation
cb: dQ
Fe 9Cr IMo 8 V,Nb
s
In 10 8
8 6 4
• •
•,
•'i
Not deformed (N aT)
0
Fotl(~ued
6
Foti~ed ( 649°C, A€t
0
(649°C, A Et = 0.5 -/0, 10000 cycles, continuously
79
= 0.5 %, cycles with 30 sec tension hold
Crepl
R~9~~,M~fb28
holn)
8
8
2
0
~ ~
0
~
~
-2
0
!
0
8
-4 0.02
0.04
0.06
0
S
0
0 0
0
~
S
0.08
q
~
Curves of dI.ldO vs q for specimens of modified Fe-9.0Cr-1.0Mo steel which have undergone various types of deformation. A magnetic field of ~28 kg was applied to the specimens during the SANS measurements. A = 0.48 nm,
Source: S. Kim, J. R. Weertman, S. Spooner, C. J. Glinka, V. Sikka and W. B. Jones, "Microstructural Evaluation of a Ferritic Stainless Steel by Small Angle Neutron Scattering," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf, Eds., American Society for Metals, Metals Park OR, 1984, P 175
197
8-1. Type 301 Stainless Steel: Scatter Band for Fatigue Crack Growth Rates K.If , ksi • in.1/2 20
60
40
80 100
10- 3 GI
i3
~ ......
10- 2
/
E E
z·
/
"t:l ......
/
/
/
I /
'"
"t:l
......
GI'
'"
/ /
..c
/
/
GI
i3
10- 4
I
.~
z·
// /
i0
... 10- 3
'" u
~
...u '" GI
:::I
'" u.. '"
.;;
/
10- 4
/
/
L-.J'-
/
/
/ /
/
/
/
/
/ /
/
/
/ /
>u '<, "t:l ......
'"
"t:l
/
/ 10-5 1/2 hard tvpe 301 24°C (75°F) 0.063 < R < 0.807
--'-_ _.1...---J'---'---'---'-...J...........L - - J
10- 6
80 Effective stress-intensity factor, K.1f = Kmax [1 - Rl o.66 7 , MPa • m 1/2 Scatter band offatigue crack growth rates of Y2·hardtype 301 stainless steel, tested at 24 0 C (75 OF), 10 Hz, andR ratios of 0.063 to 0.807 based on effective stress-intensity factor, Kerr'
Fatigue crack growth rate data reported by Walker for Y2-hard type 301 stainless steel sheet are summarized in the above graph. The data were obtained in air at room temperature over a series ofload ratios (R) from 0.063 to 0.807 at a frequency of 10 Hz. These data are based on the "effective stress intensity factor," Kerr, rather than on fj,K, to account for the effect of the range of stress ratios. Kerr is defined as follows: Kerr = K max (I - R)m where m is determined empirically and R is the load ratio (minimum load/maximum load) on cyclic loading. The crack growth rate law then becomes: da/dN= C[Kmax(l- R)my Results of fatigue crack growth rate tests on austenitic stainless steels have shown that the crack growth rate tends to increase as the R ratio is increased, when compared at given values of fj,K. If tests are made at several load ratios to determine m, then the effects of other load ratios may be estimated.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for Metals. Metals Park OH, 1982, P 114
198
8-2. Type 301 Stainles Steel: Effects of Temperature and Environment on Fatigue Crack Growth Rate tlK, ksi . in. 1/ 2 20
10- 5
Gl
U
> c.>
I"'"
--.E --'"
2'
"'"
"tl "tl
10- 6
20°C (68°F) L-T
Type 301
30
40
50
2 X 10- 7
60
Stress-intensity factor range, tl K, MPa . m 1/2 -----Annealed, tested in argon - - - - - Annealed, tested in air - - - - - Warm worked, tested in argon
Fatigue crack growth rates for type 301 stainless steel have been reviewed by Pineau and Pelloux in the temperature range from - 30 to +95 °C (-22 to +203 OF).The results, summarized in this graph, were obtained on compact specimens 7 mm (0.28 in.) thick at a cyclic frequency of 20 Hz with a sinusoidal waveform at a load ratio (R) of 0.0 1. All specimens were tested in dry argon except one series that was tested in laboratory air. For the annealed specimens tested in argon, fatigue crack growth rates at a given t::.K value increased as the temperature increased over the testing temperature range. Fatigue crack growth rates in laboratory air at 20°C (68 OF) were higher than for corresponding conditions in argon, indicating that the humidity and/ or oxygen in the air influenced the growth rates. The warm worked specimens were reduced 65% at 450 to 500°C (840 to 930 OF), resulting in a substantial increase in strength. Fatigue crack growth rates for the warm worked specimens (above) indicate that the fatigue crack propagation properties of the warm worked alloy are different from those of the annealed alloy. This effect of warm working has been observed for other austenitic stainless steels. These differences are attributed to the extent of the strain-induced transformation at the crack tip. This transformation effect would be most noticeable in type 301, because it is less stable than the other alloys in the UNS S3xxxx series. Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OR, 1982, P 113
8-3. Type 304 Stainless Steel: Effect of Temperature on Frequency-Modified Strains
AISI 304 STAINLESS STEEL o • o •
LIJ
.....J
« <.,;)
6
..
430°C 650°C 8160C
(f)
z
«
a: .....
(f)
Data of Berling and Slot for AISI 304stainless steel, showing frequency-modified elastic and plastic strains at three temperatures in air,
In contrast to most other segments of our technology, interest in the fatigue problem in the power-generation industry generally involves elevated temperature. Laboratory testing on both smooth specimens and specimens designed for crack growth is performed with temperature and frequency or strain rate as parameters. The importance offrequency or strain-rate effects is shown in this chart. These data are for solution-treated AISI 304 stainless steel subject to triangular wave shapes at equal-loading and reverse-loading strain rates. Representation of the behavior here utilizes fatigue equations known as frequency-modified fatigue equations. They describe the elastic and plastic strains versus fatigue life and include the frequency or strain rate of the test. For the present purposes they are useful in showing how increasing temperature acts to change the cyclic stress-strain response and the strainlife fatigue response of this alloy.
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979. P 13
199
200
8-4. Type 304 Stainless Steel: Fatigue Crack Growth RateAnnealed and Cold Worked ~K, ksi . in. 1/ 2
20
10
30
40
60
Cold worked 25%, tested at 427°C (800°F)~
., U
I I ~// I
10- 3
> u
-<;
E E
/
/ /
.,
I'
'"
"t:l
....,'~
...s:
s:
/
/
e
IJI,'
Cl
-"u
U
> u
'<,
.~
z'
Cold worked
25%, tested at 25°C (77°F)
10-5
"t:l -..
'"
"t:l
/it'
~
.,u ::>
Cl
'':;
'"
10-4
y/~
z' "t:l
'<,
U.
I
,1:"
I I
Annealed, tested at 427°C (800°F) -
80 100 2 X 10-4 /
10- 4
/
/
"
Annealed, tested at 25°C (77°F)
Type 304 ........
10
-'--_--'-_....L.._---'_....L.----lU
20
30
40
60
10- 6
80 100
Stress-intensity factor range, ~K, 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), 0.17 Hz, and an R ratio of O.
In some applications, type 304 stainless steel components are fabricated in the cold worked condition to improve strength properties. A comparison of fatigue crack growth rate data by Shahinian, Watson, and Smith, illustrated in this graph, shows that the high-~K crack growth rates were lower for the cold worked specimens than for the annealed specimens. Crack growth rates were higher for the specimens tested at 427°C (800 OF) than for corresponding specimens tested at room temperature.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Campbell, William w. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, p 120
8-5. Type 304 Stainless Steel: Effect of Humidity on Fatigue Crack Growth Rate CiK, ksi . in.1/2
10
20
40
60
80 100
10- 4
10- 5
~Roomair 10- 4
/
/
/
20
Type 304 25°C mOF)
30
40
Stress-intensity factor range, CiK, MPa • m 1/2
Effect of humidity on fatigue crack growth rates for type 304 stainless steel tested at room temperature, 0,17 Hz, and an R ratio of O.
The effects of humid air environments on the room temperature fatigue crack growth rates of specimens of annealed type 304stainless steel are shown in the above chart for specimens cycled at 0.17 Hz with an R ratio of zero (Shahinian, Watson, and Smith). At the lower end of the t:.K range, fatigue crack growth rates in humid air are substantially greater than crack growth rates in dry air. However, fatigue crack growth rates of specimens oftype 304stainless steel tested in a pressurized water reactor environment at 260 to 315 °C (500 to 600 °F) with R ratios of O. 2 and 0.7 were no greater than the fatigue crack growth rates in air at the same temperature with an R ratio less than 0.1 (Bamford). However, variations in R ratios influenced the fatigue crack growth rates in the pressurized water reactor environment.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH. 1982. p 122
201
202
8-6. Type 304 Stainless Steel: Effect of Aging on Fatigue Crack Growth Rate l1K, ksi . in.1/2
10-4
10- 3
.,
u
> lJ
-.~
10- 5 Z ~
'"
"t:J
10- 4
Unaged
Aged
• •
o
A
Hold time Zero 0.1 min 1.0 min 10- 6 Type 304 593°C (1100°F)
10-5
.l--_ _....L..._--'-_....I....----I_L-.L-J........J
L-..JL-.I--'-
10
20
40
60
80
100
Stress-intensity factor range, l1K, MPa . m 1/2 Effect ofaging at 593 ° C (1100 OF)for 5000 h, and hold times of 0.1 and 1.0 min for each cycle, on fatigue crack growth rates of L-T oriented specimens of type 304 stainless steel tested in air at 0.17 Hz and an R ratio of O.
Because austenitic stainless steels are expected to give long service life, an evaluation of the effect oflong-time aging at service temperatures is important. Results offatigue crack growth rate tests on specimens that were tested in the unaged and aged conditions (5000 hours at 593 °C, or 1100 OF) are shown in this graph, as reported by Michel and Smith. After aging for 5000 hours at this temperature, precipitation of M 23C 6 carbides is essentially complete. These results indicate that at 593 °C (1l00 OF) there are no deleterious effects of aging on the crack growth rates of specimens that are continuously cycled. When a holding time of 0.1 or 1.0 minute is included in each loading cycle, there tends to be a slight increase in the fatigue crack growth rate at a given 11Klevel.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 121
8-7. Type 304 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate .1K, ksi . in,1!z
10
20
649O C l 1 2
.......
E E
10-3
z·
y
10-4
t<, 1/ .:
"tJi'"
..
E
~
.'" "". '"
100
538"C (1000"-><-/
~
i0
60
,
II>
c:;
>u
40
10- 4
f!
u u
II> :::l
t'I~
II>
10- 5
'"
u,
10- 5
"'" "
.......
~~~
10-6
316°C 1600°F)
10
>-
u .......
.E z·
~'
...'"
c:;
20
Type 304
40
60
100
Stress-intensity factor range, .1K, MPa . m l/Z Effect of testing temperature on fatigue crack growth rates for annealed type 304stainless steel tested in air at 0.066 Hz and anR ratio of 0 to 0.05.
Results offatigue crack growth rate tests on types 304 and 304L stainless steel at room temperature and at elevated temperatures have been reported by James and Schwenk, and by others. As shown in this graph, increasing the exposure temperature from room temperature to 650 °C (1200 ° F) increases the fatigue crack growth rates at any ~Klevel within the range ofthe tests in an air environment. These data, reported by James and Schwenk, are for specimens of both the L-T and T-L orientations, for several different maximum alternating loads, for load ratios of 0 to 0.05, and for cyclic frequencies from 0.033 to 6.66 Hz for the room-temperature tests and 0.067 Hz for the elevated-temperature tests.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park DB, 1982, P 115
203
204
8-8. Type 304 Stainless Steel: Damage Relation at 650°C (1200 OF)
o
o
STRAIN RATE, IN.lIN.lSEC 4 x 10'3
o (). ~
'i7
HOLDPERIOD, MINUTES TENSION COMPRESSION o 0 I 0 10 0 30 0 60 0
180
Ll
o
Q
o ~
(3
0 30
3 30 30 0.1 0 0
3 30 3 0 0 0
I
-tI
o
o
4 x 10'3 4 x 10'4 4 X 10'5
O"rtl€pN,
I
I3 v l3 ( k. I ) = C = 1.158
C
x 10 5
13 = 0.895 k
= 0.756
101' - : - - -.........- -.........- - ' -.........L.;:-----''------''-----'---'--'-=----'----'----'---'---'-:-----'----'--.........- I -..... 101 105
Ostergren's damage relation for AISI 304 at 650°C (1200 OF).
The damage function was proposed by Ostergren and is based on the frequencymodified fatigue approach. A damage function is approximated by the quantity U,!:iE p , where u, is the maximum stress in the cycle and !:iE p is the inelastic strain range. The tensile hysteresis energy is employed to account for the facts that low-cycle fatigue is essentially a crack-growth process and that crack growth and damage occur only during the tensile part of the cycle. The use of the tensile-stress quantity, in conjunction with the plasticstrain range, provides a means of accounting for loop unbalance, since, for the same inelastic strain, a positive mean stress provides a greater hysteresis energy than does a compressive mean stress. The method is effective in accounting for hold-time effects, as indicated in the chart above.
Source: L. F. Coffin. "Fatigue in Machines and Structures-Power Generation, "in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 23
8-9. Type 304 Stainless Steel: Fatigue Crack Growth Rate at Room and Subzero Temperatures
5
10-4
10-3 1-- +-- - +-- - - - - - +-'''''10,......, • 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, 11K, MPa
100
vm
Fatigue crack growth rate data for type 304 austenitic stainless steel (annealed) at room temperature and at subzero temperatures. For this alloy, crack growth rates are nearly the same at room and cryogenic temperatures.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, p 756
205
206
8-10. Types 304 and 304L Stainless Steel: Effect of Cryogenic Temperatures on Fatigue Crack Growth Rate ~K. ksi • in. 1/ 2
5
20
40
60
80
Gl
U
~ ...... E E
10- 3
Z
"C ...... Cll
Gl
U
"C
~i
Type 304L 22°C (72°FI
E
...i!=
10- 5
~
z
--
Cll
"C
tJ)
.>t!.
u
...u'" ~
tJ)
-. .E "C
...0
Gl
> u
10- 4
Type 304L -196, -269°C (-320, -452°FI
.;;
'"
IL
10- 6
Stress-intensity factor range, ~K, MPa • m 1/2 Fatigue crack growth rates for annealed types 304 and 304L stainless steel at room and cryogenic temperatures, 20 to 28 Hz, and an R ratio of 0.1.
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, or -452 OF)are shown in this graph. The data for type 304 were scattered over the range shown, while for type 304L, the data at room temperature described one curve and the data at the cryogenic temperatures described the other curve. 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. Furthermore, if design calculations for type 304L are based on room temperature fatigue crack growth rates, the calculations will be conservative for cryogenic exposure.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 123
8-11. Type 304 Stainless Steel: Fatigue Crack Growth Rate in Air With Variation in Waveforms ~K,
10
ksi . in.
2
'/ 30
20
40
50
10- 4 CIl
U ~ .......
E E
10- 3
/ / / / / I / / / II / / Waveforms
Z· '1::J ....... co '1::J
2l'
e
s:
i0
.. ..
( ( / / / /
."
~
u co u
10- 4
/ / / / / /
CIl :J
." .;;
co
u,
rYYY\
u
10- 5
20
30
40
.E Z·
} 0.067 Hz
Type 304 538°C (1 0000 F) R = 0.05
/
>-
~ co '1::J
/VVV\
/
10
.!! u .......
10- 6
50
Stress-intensity factor range, ~J<, MPa . 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 0.05 with two different waveforms at 0.067 Hz.
The data presented in this graph were obtained in tests with a sawtooth waveform. Changing from a sawtooth waveform to a waveform with a short holding period at maximum load did not influence the overall fatigue crack growth rates according to additional data reported by James and shown above.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society For Metals, Metals Park OH, 1982,P 117
207
208
8-12. Type 304 Stainless Steel: Effect of Hold Time on Cycles to Failure
30
LABORATORY TESTS AISI304 STAINLESS STEEL 650°C
~
z ......
180
Z ILl C>
z 0.01 « a:::
z «
60
o
a::: ~
en -J
~
g
60
DO 01
10 0
30 30
o I
0
I
o NO HOLD TIME o TENSILE HOLD TIME IN MINUTESAS INDICATED
0.001 L...-_--L._ _.L..---L----L---l.-_ _.L..-_----L_..L-...l..-.L..-_-..l..._ _--'--~ 100 1000 10,000 LIFE - CYCLES TO FAILURE Effect of hold time on life for AISI 304 stainless steel.
Wave-shape effects are also important in fatigue crack growth, as has been studied by Barsom. He observed that the crack growth rates were greater as the loading rate increased and the unloading rate decreased, given a fixed period of cycling. Overload effects are also important in retarding crack growth. Substantial damage can result from these wave shapes, particularly when the hysteresis loop is severely unbalanced, as can occur in long tensile-strain hold-time tests.
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 19
8-13. Type 304 Stainless Steel: Effect of Hold Time and Continuous Cycling on Fatigue Crack Growth Rates 6K, ksi . in. 1/ 2
10
40
20
60
80 100 10- 1
.:E.
E
E
I
I
0.1 min hOld-...f I
J
I
"
"
~
/
,
,r~
10- 2
1.0minhold
Type 304 593°C (1100°F)
10- 1 '--_ _---'_ _....L-_...L-_---'_---'_...L...I 20 30 40 60 80 100 Stress-intensity factor range, 6K, MPa . m 1/2 Fatigue crack growth rates per unit of time ida]dt) for annealed type 304 stainless steel for continuous cycling (0.17 Hz), for 0.1 and 1.0-min hold times at maximum load for each cycle at 593 ° C (1100 OF), and for anR ratio of O.
As shown above, the fatigue crack growth rate is greater for specimens tested with no holding time (continuous cycling) than for specimens held at maximum load for 0.1 or 1.0 minute per cycle. The lowest fatigue crack growth rates occurred for specimens with the longest holding time, based on dal dt, The same trend was observed for tests at 593°C (liOO OF), as shown here. Therefore, cyclic loading has a more damaging effect than static loading on crack growth per unit of time.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 118
209
210
8-14. Type 304 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate dK, ksi . in.1/ 2 10
20
30
0.0014 Hz
40
50
0.0067 Hz
10- 4
Q)
U
>o
--EE z' --'"
10-3
"tl "tl
l!l'
Q)
u
~
...::
10- 5
s:
een
"tl
..10:
u
~
"tl
10- 4
u
Q)
>u
--.5 z' --'"
;:,
en
'':;
'"
u..
Type 304 538°C (lOOO°F) R = 0.05
10
20
30
40
10- 6
50
Stress-intensity factor range, d K, MPa . 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 0.05 in air with a sawtooth waveform.
For fatigue crack growth rate tests on specimens of annealed type 304 stainless steel at elevated temperatures, increasing the cyclic frequency will decrease the crack growth rate over part of the ~K range, as shown here for tests at 538°C (1000 OF).
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William w. Gerberich and John H. Underwood. Eds., American Society for Metals, Metals Park OH, 1982, P 116
8-15. Type 304 Stainless Steel: Effect of Frequency on Fatigue Crack Growth Behavior
~K,
SiRESS INTENSITY FACTOR RA:\iGE,
kg/lmm?/2
Il
5x 101
.., ~
.l::!
s: u
c:
z· ::!2
"'
't:l
....-c. ~
co:
10-5
....
:I:
;: 0
co:
<.:l
:><:
u
e:(
co:
ANNEALED TYPE 304 S S. TESTED IN AIR AT 53S"C 1l000°f) R· 0.05, Ref. [45]
y
~
:::I <.:l
.... -e
+
......
0.003 cpm
'il 0.4 cpm
00 4cpm
10-6
o
40cpm
li. 400 cpm
o
4000cpm
4 10
STRESS INTENSITY FACTOR RANGE,
~K, Ib/lin)3/2 5x101
STRESS INTENSITY FACTOR RANGE, oM, 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).
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 14
211
212
8-16. Type 304 Stainless Steel Welded With Type 308: Fatigue Crack Growth Rates t.K. ksi . in. 1I2 t.K. ksi .
in. 1/ 2
40
20
20
40
60 80
60 80
10- 2
Ql
10- 4
Ql
c:;
~
E
10- 4
--
E E
Shielded metal arc
E Z
c:; > u
z·
--. 10... "t:l
10- 3
"t:l
"t:l
co -e
Ql
c:; > u
...eoi
j!l'
--.S ...::.. z· '" --.. e
i0
"t:l
~
10- 5
~
"t:l
s:
u u
Submerged arc
~
..'"
u
.;:;
~ .;:;
u..
10- 4
--.S z· --.. "t:l
~
Ql
..~
•
Ql
c:; > u
0
s:
u..
Shielded metal arc
3
10- 5
"t:l
10- 4 o SMAW 1 • SMAW 2
24°C (75°F) 20
40
10-6 60 80 100
Stress-intensity factor range, t.K, MPa . m 1/2
593°C (l100°F) 20
40
10-6 60 80 100
Stress-intensity factor range, t.K, MPa . 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), 0.17 Hz, and an R ratio of O.
Type 308 stainless steel is the alloy that is usually used for welding rod for weldments in type 304 stainless steel when those weldments are to be exposed to room temperature or to elevated temperatures in service. Because service experience has shown that failures are more likely to originate in weld metal or in heat affected zones than in the base metal, it is important to have fracture information on weldments. In general, fatigue studies at elevated temperatures on specimens from type 304 weldments have shown that the fatigue crack growth rates in the type 308 weld metal and heat affected zones are no greater than in comparable specimens of the base metal. Fatigue crack growth rate data obtained by Shahinian for specimens of type 304 welded with type 308 rod by the submerged arc and shielded metal arc processes are shown above for tests at room temperature and at 593°C (llOO°F).
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood. Eds.. American Society for Metals. Metals Park OH. 1982. P 125
8-17. Types 304 and 310 Stainless Steel: Effect of Direction on S-N 400 Type 310, transverse Type 310, longitudinal
Q.
~" 300 ~
~E 200 :::l
'iii
..><:
40 '" .......'" 30 '"
","
'"...
,~
50
~~-~
Type 304, longitudinal
E :::l
20 ,~ x
~ 100
~
10 105
106 No. of stress cycles
"'
~
0 108
.SoN curves for two grades of stainless steel,
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 32
213
214
8-18. Types 304, 316, 321, and 348 Stainless Steel: Effects of Temperature on Fatigue Crack Growth Rates AK, ksi . in. 1/ 2
20 10 60 100 40 10-2 ...-r----r----.------r---.---. 4 X 10-4
10- 4 Q)
u
> Ll
--EE z· --'a>" "0
10- 3
"0
.!!!
... s: ...3:
Ll
> Ll
~
-z· -.~
e
"0
Cl
~
l'Il
..
"0
Ll
l'Il
10-5
Ll
Q)
::::I
Cl
.~
l'Il
IL.
10-4
L -_ _....L.
10
20
10-6 80 100
....I...._--l._...l.-..J.J
40
60
Stress-intensity factor range, AK, MPa • m 1/2
Fatigue crack growth rates for annealed types 304, 316, 321, and 348 stainless steel in air at room temperature and 593°C (1100 OF), L-T orientation, 0.17 Hz, and an R ratio of O.
As reported by Shahinian, Smith, and Watson, fatigue crack growth rate tests were made on singleedge-notch cantilever specimens oftypes 321 and 348 stainless steel from L-T orientation at 0.17 Hz with an R ratio of zero at room temperature and at elevated temperatures to 593°C (1100 OF). As for types 304 and 3I6, fatigue crack growth rates in air increased with increasing testing temperature. The curves above show that, at room temperature, the fatigue crack growth rates for types 304, 316,321, and 348 all fall within a narrow band. For tests at 593°C (1100 OF),however, specimens of type 3I6 had the least fatigue crack propagation resistance, whereas specimens of type 348 had the highest fatigue crack propagation resistance, over the 11Krange studied. Results of tests on specimens of types 304 and 321 were nearly the same at 593°C (1100 OF) in air. Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 138
8-19. Type 309S Stainless Steel: Effect of Grain Size on Fatigue Crack Growth Rate ~K, ksi •
10
20
40
in.1/2
100
r - - , - - - r - - - , - - - r - - - - - - - - - - , 10- 3
10- 2
10-4
Type 309S Testing frequency, Hz 10 15 20 25 30
Grain size 45 fJm 480 fJm -0-
-0-
10- 6
10- 7 10
20
40
100
Stress-intensity factor range, ~K, MPa • m 1/2 Fatigue crack growth rates for annealed type 309S stainless steel for two grain sizes, at frequencies from 10 to 30 Hz and an R ratio of 0.05 at room temperature in air.
Types 309S and 3 lOS stainless steel are the low-carbon versions of types 309 and 310. They have higher chromium and nickel contents than type 304 and consequently have better corrosion resistance and more stable austenite than type 304. Fatigue crack growth rate data have been reported by Thompson for tests made at room temperature on compact specimens from plate of type 309S and the L-T orientation after heat treating to a grain size of 45 Jlm in one set and 480 Jlm in the second set. Specimens with the smaller grain size had substantially higher yield and ultimate tensile strengths than the specimens with the larger grain size. Fatigue crack growth rates were obtained on tension-tension loading at frequencies from 10 to 30 Hz and at an R ratio of 0.05. The results are plotted above. These data provide further evidence that a wide variation in grain size, and the associated variation in strength level, does not affect the results of fatigue crack growth rate tests.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for Metals, Metals Park OH, 1982. P 126
215
216
8-20. Type 31 OS Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate ~K, ksi • in. 1/ 2
20
40
.-------r---r-..---..---.......,."""T""'l~---.______.
., U
> ("l
10-4
10- 3
--
E E
z·
Base metal -196, -269°C (-320, -452°F)
--.,' "C
co
"C
......co
.,
u
> ("l
10- 5
...
.c
--.5 z· -"C
~
een
co
"C
-"o co ...
., ("l
10- 4
:::l
en '':; co u..
Type 3105
20
40
60
100
Stress-intensity factor range, ~K. MPa • m 1/2 Fatigue crack growth rates for annealed type 31OSstainless steel at 22, -196, and - 269 0 C (72, - 320, and -452 0 F), 10 to 28 Hz, and anR ratio of 0.1, with corresponding data for SMAW welds with type 316 filler metal.
Because of its high nickel content, type 3lOS stainless steel is completely stable at all cryogenic temperatures and with any amount of cold working. Therefore, it is often considered for cryogenic applications that require a high degree of austenite stability on thermal cycling and strain cycling. Fatigue crack growth rate at various temperatures is illustrated above.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for Metals, Metals Park OH, 1982, P 127
8-21. Type 316 Stainless Steel: Growth Rate of Fatigue Cracks in Weldments liK, ksi • in. 1/ 2
10-2
20
40
60
100
4 X 10-4
Ql
10-4
13
> u
'<,
E E
z' ~
<0
10- 3
Ql
13
> u
"C
'<,
fl' E
.~
z' "C
s:
~
..
'<,
<0
0
"C
Cl
10- 5
~
u
E u Ql
:::J Cl
'.J
<0
u..
10- 4
Type 316 593°C (1100°F)
10- 6 20
40
60
100
Stress-intensity factor range, liK, MPa . 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 1.2 X 1022 n] cm-, >0.1 MeV at 410°C (770 OF)].
Results of fatigue crack growth rate tests on weldments of type 316 stainless steel have shown that the crack growth rates in the weld metal are generally no higher than in the base metal and may be somewhat lower at elevated temperatures (Shahinian, Smith, and Hawthorne). The curve shown above for unirradiated weld metal tested at 593°C (1100 OF) represents fatigue crack growth rates substantially lower than those for the unirradiated base metal at any given I:>.K level (Shahinian). The weld was produced by the submerged arc method using type 316 welding rod. Weldments were stress-relief annealed at 482°C (900 OF). Specimens were single-edge-notch specimens for cantilever loading and were tested at 0.17 Hz and at an R ratio of zero. Irradiation slightly reduced the fatigue crack growth resistance of the weld metal, but its fatigue crack growth resistance was better than that of the unirradiated base metal.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds .. American Society for Metals, Metals Park OH, 1982, p 134
217
218
8-22. Type 316 Stainless Steel: Fatigue Crack Growth RatesAged vs Unaged AK. ksi . in.1/2
10 60 80 20 40 10 2 ."......---r------,;-------r--,--..,
Type 316, cold worked, 0 0 tested at 593 C (1100 F )
III
U
e
EE
10- 2
z ~ "lJ
III
10- 1
U
10- 2
--.S --..
> u
J!l' E
..c:
10- 3
i
e ;
Z
"lJ
"lJ
u
E u
III :J
10-4
..
en .;;
u..
10- 5 10- 4 0
Ii.
10- 5
v L-_....L-
10
--l.
20
Aged Hold time Zero 0.1 min 1.0 min
• • •
10.0 min
10- 6 80 100
...L-_--l._......L~
40
60
Stress- intensity factor range, AK. MPa . m 1/2 Effect of exposure at 593 °C (1100 OF) for 5000 h, and hold times during cycling, on fatigue crack growth rate of20% cold worked type 316 stainless steel at 593 ° C in air.
Results also have been reported by James for fatigue crack growth rate tests in 20% cold worked specimens of type 316 stainless steel which were cycled at frequencies of 0.0055 to 6.66 Hz, at 538°C (lIDO OF) and at an R ratio of 0.05. Over the 11Krange studied, the fatigue crack growth rates were highest for the specimens subjected to the lowest cyclic frequency.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Material, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for Metals, Metals Park OH, 1982, p 133
8-23. Type 316 Stainless Steel: Fatigue Crack Growth RatesEffect of Aging .1K, ksi • in. 112
20 40 60 80 10 10-1.......- --r-----.-------,----.--,....,
10-3
10-2 Q)
U
-.~
593°C (1100°F)
E E
z·
10- 4
~
'"
Q)
"tl
...e
U u>'<,
Q)'
..c
10-3
.~
i
z·
een
"tl
'<,
'"
"tl
~
u u
~
10-5
Q)
:::l
en ';:;
'"
Type 316 593°C (1100°F)
LL.
10-4
Unaged Aged
•
0
•...
G
A
Hold time Zero 0.1 min 1.0 min
10-6
10-5 10
20
40
60
80 100
Stress-intensity factor range, .1K, MPa . m 1/2 Effect of exposure in air at 593°C (1100 OF) for 5000 h, and hold times, on fatigue crack growth rates for annealed type 316 stainless steel at 593°C in air.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, p 130
219
220
8-24. Type 316 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate dK, ksi • in. 1/ 2
10
20
40
60
100
10- 4
10- 3 Ql
1)
>u
10- 5
--.5 --
2"
"'C III "'C
10-4
10-6 Type 316 Cold worked
10- 5
10
20
40
60
100
Stress-intensity factor range, dK, MPa . m 1/2 Fatigue crack growth rates of20% cold worked type 316stainless steel for various temperatures. Curves are averages for L-T and T·L specimens at each temperature in air; 3 Hz at 24°C, 0.67 Hz at elevated temperatures; R = 0.05.
Results of tests on compact specimens of 20% cold worked type 316 stainless steel at frequencies of 0.67 and 3.0 Hz and at an R ratio of 0.05 are summarized in the above graph. Similar results have been reported by Shahinian for tests on cold worked type 316 at 427°C (800 OF).
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 132
8-25. Type 316 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate dK. ksi • in. 1I2
10
20
30
40
50
60 10- 4
0.0067 Hz----... II>
U
I
10- 3
> o
--
)7 ~0.67"'
E E
z· ~
'"
0067
"C
...eai
...;: e
"'-y' //J '~
II>
10- 5
f
s:
Cl
"'..u'""
1,1
10- 4
U
> u
--.5
z· "C
<,
'"
"C
U II>
::>
Cl .;;
'"
u,
Type 316 538°C (lOOO°F)
10- 6
10 Stress-intensity factor range, dK, MPa • m 1/2 Effect of variation in cyclic frequency on fatigue crack growth rate of annealed type 316 stainless steel in air at 538°C (1000 OF) and an R ratio
of 0.05.
As may be observed above in tests at frequencies in the range from 0.0067 to 6.67 Hz at 538°C (1000 OF), the trend is for the crack growth rate to increase as the frequency is decreased, but there is more scatter than for type 304. In studying heat-to-heat variations in fatigue crack growth rates for specimens from three heats of type 316stainless steel, 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 2.6 over the range of 11K values studied. One heat was produced by air melting, another by vacuum arc remelting, and a third by double vacuum melting.
Source: J. E. Campbell, "Fracture Properties or Wrought Stainless Steels," in Application or Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood. Eds .• American Society for Metals, Metals Park OH, 1982, P 131
221
222
8-26. Type 316 Stainless Steel: Fatigue Crack Growth Rate in the Annealed Condition 6K, ksi . in. 1I2
10
20
30
40
60
80 100 10- 3
/1
370·C (700.F)
---t!
10- 2
""1/;/ I ,
482·C {900·
/1
,1
p'"
//' fi,' I :,/ 593·C .' (1100·F) /
l'
10- 4
/
..
u
>-
"I
~
y: /:
.5
z· ~
./ ;,'I
:i
"
I'"I
/ /'1
10- 5
I
10- 4
316 ' - -_ _'-_J.----'_L-l......J-J....I-J..J
10
20
30
40
60
10- 6
80 100
Stress-intensity factor range, 6K, MPa • m 112
Effect of testing temperature on fatigue crack growth rates for annealed type 316 stainless steel tested in air at 0.17 Hz and an R ratio of O.
Most of the fa tigue crack growth rate testing on type 316 stainless steel has been oriented toward its use in components for nuclear reactors, but the data also are applicable to design of equipment for fossil fuel power stations, petrochemical refineries, and chemical plants. Its improved yield strength compared with that of type 304 stainless steel is an advantage for these applications. The austenite stability in type 316 is greater than that in type 304, so it is advantageous to use type 316 rather than type 304 for critical cryogenic applications. Effects of elevated temperature on crack growth rate are summarized in the graph above.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 129
8-27. Type 316 Stainless Steel: Effect of Environment (Sodium, Helium, and Air) on Cycles to Failure 10.0 8.0 DATA POINTS
6.0
o
4,0
x
a
FATIGUE TEST MEDIA Na (10 ppm OF 02) AIR HELIUM
t'
z
2.0
~
ex:
l-
V>
u
1.0
...J u 0.8 >u 0.6
0.4
EXPOSED SPECIMENS • FATIGUED IN SODIUM FATIQJEDIN AIR FATIGUED IN HELIUM
EXPOSURE MEDI~
•
•
0.2
Na
ue ppm OF 02)
EXPOSURE CONDITION 286 hrs AT 92S K
0.1 102
3
4
6
8 103
2
6
8 104
CYCLES TO FAILURE
Effect of environment on fatigue characteristics of type 316 stainless steel at 92SK; based on cyclic strain and cycles to failure.
Source: R, H. Cook and R. P, Skelton, "Environment-Dependence of the Mechanical Properties of Metals at High Temperature," in Source Book on Materials for Elevated-Temperature Applications, Elihu F, Bradley, Ed" American Society for Metals, Metals Park OH, 1979, P 84
223
224
8-28. Types 316 and 321 Stainless Steel: Effects of Gaseous Environments on Fatigue Crack Growth Rates aK, ksi . in.1I2
8 10 20 10-3 r----,,-..----....
40 60 100 ---...,.:.-......:.=----..:..;
Type 316 649°C (1200°F)
">
U
10-5
Roomair
u
E E
z· ~ ."
Type 316
2SoC 177°F)
10-'
Dry air Wet nitrogen Dry nitrogen
fi
f! .c
"
U
> -!:! .~
~
z·
12tn
~
.><
<0
."
u
f!
u
":tn>
.~
10-6
Types 316 and 321
u,
2SoC (77°F) Room air
Wet air
10- 5
Types 316 and 321 649°C (1200°F) Dry nitrogen Dry argon
8
10
20
40
60
100
Stress.intensity factor range, 6K, 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).
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. Fatigue crack growth rates in room air at room temperature were the same for types 316 and 321 stainless steel. Furthermore, in tests at 649 °C (1200 OF)in dry nitrogen, fatigue crack growth rates for types 316 and 321 also were the same. In air, however, fatigue crack growth rates in type 316 specimens increased by a factor of about 22 over rates in an inert environment at the same temperature.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, p 135
8-29. Type 321 Stainless Steel: Effect of Hold Time on Fatigue Crack Growth Rates .6K, ksi . in. 1/ 2
10
10-'
20
40
60
80
10- 3
Type 321 593°C (1100°F)
10- 2
..
U > u
10-'
E E
z·
. "
~
.
10- 3
u
>
B
~
~
.~
~
e '" ""eu
..
10- 5
z·
..
~
"
10-'
u
::I
'"
.~
u,
10- 6
Unaged Aged 10- 5
0
•
•...
Hold time Zero 0.1 min 1.0min 10-
10- 6
'--_-'-
10
..J....
20
'
.......L_ _.l----JL-..J
40
60
80 100
Stress-intensitv factor range,ll.K, MPa . 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 (0.17 Hz), with 0.1 and 1.0-min hold time at anR ratio of 0 at 593°C (1100 OF).
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. Aged specimens tested with zero holding time had lower crack growth rates than corresponding specimens that were not aged (see above graph). Fatigue cycling with holding times of 0.1 and I,D minute on each cycle increased the crack growth rates slightly, as shown in the figure,
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics ForSelection of Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H, Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 139
225
226
8-30. Type 403 Stainless Steel: Effect of Environment on Fatigue Crack Growth Rate ~K,
ksi • in. 1/ 2
10 20 30 40 60 80 10-3 rr----,-----,--,--...........-,------,
I /
I
/
II
/ 10-4
/
/
/
10-5
Air
Type 403
z
~
In H20
III
"1:J
pH 7, 25°C pH 10, 25°C -- - -
pH 7, 100°C pH 10, 100°C 10- 6
In 1M NaCI solution
- - - - pH2tol0, 100°C 10- 5 L..-_ _---L_ _L.--L---L----I---L...J......L.J. 10 20 30 40 60 80 100
--'
Stress-intensity factor range, ~K, MPa . m 1/2 Fatigue crack growth rates in type 403 stainless steelin air, water, and aIM NaCI solution at 10 Hz and an R ratio of 0.5.
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, as shown on a different scale in the above graph. Tests in the 0.01 M (molar) and 1.0 M sodium chloride solutions were made with the solutions at pH levels of 2, 7, and 10 and with an open circuit. Fatigue crack growth rates in 0.01 M sodium chloride at pH 10 and 100°C were the same as those in water at 100 "C. At lower cyclic frequencies, the fatigue crack growth rates were higher than at 40 Hz at 6.K values above 20 MPa· m 1/2 (18 ksi· in. 1/2). For tests in the 1.0 M sodium chloride solution at 100°C (212 OF) (see graph), fatigue crack growth rates were the same as for water at the same temperature. At 100 °C (212 OF), fatigue crack growth rates in 1.0 M sodium phosphate solution at pH 10 and at 10 and 40 Hz and in 1.0 M sodium silicate at pH 10 and at 10 Hz were practically the same as those in air.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Campbell, William W. Gerberich and John H. Underwood. Eds., American Society for Metals. Metals Park OH, 1982. P 147
8-31. Type 403 Modified Stainless Steel: Scatter of Fatigue Crack Growth Rates
'--_-'-_-'-......L.......L.....L....w....LJ..._ _- ' - _ - ' -
20
40
60
100
----I
10-6
200 300
Stress-intensity factor range, ll. K, MPa . m 1/2 - - - - 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 8.3 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, tested at 10 Hz and an R ratio of 0.083 or 0.067.
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. Furthermore, exposure at 27 1°C (520 OF)in distilled water at a pressure of 8.3 MPa (1200 psi) increased the fatigue crack growth rate.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 145
227
228
8-32. Type 422 Stainless Steel: Fatigue Crack Growth Rates in Precracked Specimens Koff' ksi • in. 1/2
10
20
40
60 80 100
200
71°e (160°F) 57°e (135°F)
I
a>
10-
4
c
.:.::::
z
~
~ "tl
Type 422
~_---l
20
---L_---l_....L.........L.
l..-
..J
10-5
200
Fatigue crack growth rates in precracked round rotating beam specimens of type 422 stainless steel in 4.5% NaCl solution at room and elevated temperatues, 10Hz, and an R ratio of-I.
Type 422 stainless steel contains nickel, molybdenum, and tungsten, as well as 12%chromium to improve properties. 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. Calculations for maximum stress-intensity factors were based on equations for solid round bars subjected to bending loads. The material for these tests apparently had been heat treated to a yield strength of approximately 827 MPa (120 ksi). The specimens were one inch in diameter in the test sections. 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. Tests with several concentrations of salt solution indicated that the maximum corrosive effect was obtained with the 4.5% solution. Results oftests with specimens in the 4.5%sodium chloride solution at room temperature, 57°C (135 OF), and 71°C (160 OF) are shown above. Increasing the temperature of the solution substantially increased the fatigue crack 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 150
8-33. Type 422 Stainless Steel: Fatigue Strength-Longitudinal vs Transverse CLASS II (Crucible 422) turbine-wheel forgings, 12 to 30 inches in diameter, ksi 1900 F (l 040 C)oil quench + 1200/1400 F(650/760 C) 100..--------------------------. UNNOTCHED
x 80-
o
•
A
60r-
o
1-----------... VIBRATING NOTCHED 401-
Kt = 2.1 60· notch 0.030·inch root radius
x
CANTILEVER TESTS FOR 108 CYCLES
o
x 0% delta ferrite
.x o
• 5% delta ferrite 15/16% delta ferrite A 20% delta ferrite
o
o
A
20 .....
I 20
I
I
I
40 60 80 TRANSVERSE FATIGUE STRENGTH
100 ksi
Transverse fatigue strength as related to longitudinal fatigue strength for type 422 stainless steel, including effects of varying amounts of delta ferrite.
Source: J. Z. Briggs and T. D. Parker, "The Super 12%Cr Steels," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed.. American Society for Metals, Metals Park OH, 1979, p 121
229
230
8-34. Type 422 Stainless Steel: Effect of Temperature on Fatigue Strength CLASS II (Crucible 422) (VacuumMelted)
ksi 140
15% delta ferrite 1800 F (980 C)oil quench + tempered to a tensile strength of 131/138 ksi
130 f120 ,....(/) (/)
110 ,....-
IJJ a:: 100 "-
J-
(/)
80 70 90
601()~
'.,""'-ee_ ....... •
room temperature
-.-
700 F(370 C)
...
•
ROTATING CANTILEVER-BEAM TESTS I 105
I
I
106 CYCLES
101
108
SON curves for vacuum-melted type 422 stainless steel with 15% delta ferrite, showing effect of temperature on fatigue strength.
Source: J. Z. Briggs and T. D. Parker, "The Super 12% Cr Steels," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, P 121
8-35. Type 422 Stainless Steel: Effects of Delta Ferrite on Fatigue Strength CLASS II (Crucible 422) 3/4 -inch-diameter bar stock or 3/4 -inch-thick plate 1900 F (1 040 C)oil quench ksi 110'
100
en en LLI
0: I-
90
en
\
% DeltaFerrite TensileStrength, ksi 155 160 .... 140 5% Ferrite,Tempered 1150 F (620 C)
~ 15~20 '~~v_
~... 00% Ferrite,Tempered 1150 F
--.. . . . . .-v..
-~ "-.. 0 (620 C) o 0.: 15/20% Ferrite,Tempered 1200 F(650 C)
0-
80
V;;
a
lr-+
0-
'LONGITUDINAL ELECTROMAGNETIC CANTILEVER-BEAM TESTS
70
lOS
107 CYCLES
S-N curves for type 422 stainless steel, which demonstrate the adverse effects of
delta ferrite on fatigue strength.
Source: J. Z. Briggs and T. D. Parker. "The Super 12%Cr Steels," in Source Book on Materials (or Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH. 1979, p 121
231
232
8-36. 17-4 PH Stainless Steel: Fatigue Crack Growth Rates in Air vs Salt Solution .o.K, ksi • in.'/2 10
100
10-2 I
'---Hll00 , R =0.05
'I
1 min hold Salt soln
,~ I~
10-4
I I
10- 3
il!
'" ~u "5,
10- 4
."u..
10-6
10- 5 17·4 PH '--
-'-
-'-
10
100
...1..1
10-1
Stress-intensity lactor range, .o.K, 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 3.5% NaCI solution,
Results of fatigue crack growth rate tests on specimens of 17-4PH stainless steel under comparable conditions are presented here. Those specimens that were tested in the HI 050 condition at a stress ratio of 0.67 with a one-minute holding period at maximum load in each cycle had the highest fatigue crack growth rates (as for 15-5PH) in the upper levels of I:!.Kvalues. Specimens in the H II 00 condition tested in a salt solution with a one-minute holding period, however, had fatigue crack growth rates only slightly higher than those of comparable specimens tested in air with continuous cycling.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals. Metals Park OH. 1982, P 156
8-37. 15-5 PH Stainless Steel: Fatigue Crack Growth Rates in Air vs Salt Solution LlK, ksi • in. 1/2 10
100
) "
Hl~50 ~I //'I R - 0.67 I:/, 1 min hold I 1//
.
I/Il
RT.ir
,,1
l,' l'
'li
~
E
10- 3
E
1
z·
..
Hll00--.......,' R = 0.05 ~ 1 min hold " I I 'I Salt soln
I
~
."
!l'
e
VI 'I1 ,r--HllOO R = 0.05
.t::
i
e '" ""ut!
.. .§.
---Hl050 R = 0.05 10Hz Sine wave RT air
10-'
~
u
I'
,
TO
1 min hold RT air
I
u,
10- 6
H 1100 :----""'/' R = 0.05 10 Hz I Sine wave RTair
I
10-5
15·5 PH 1-
--L.
...1...
10
100
.J1O-
7
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
233
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
> u E E
---
10- 4
Z
-e
Gl
l'O
U
.....
"tl
> U
---
Gl'
..
.~
l'O
.s:
10- 3
Z'
;:
"tl
e
l'O "tJ
Cl
~
u
E u
Gl
:::l
Cl
';; l'O u..
10-5
PH 13-8 Mo H1100
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
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
235
236
8-40. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Subzero Temperatures ~K,
6
4
8 10
ksi . in. 1/ 2
20
40
60
100 10- 4
10- 5
Q)
u
> u '-: .!:
z
1:1 "nl 1:1
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
237
8-41. PH 13-8 Mo Stainless Steel: Constant-Life Fatigue Diagram Minimum stress, ksi -150
o
-100
100
+50
150
200
1600 1400
200
1200
'" a.. :2 1000
150
Jl
100
.S x '" :2
",'
...~'"'" E E
800
S)<::l l'l'b
'x
'" :2
E
'li~
:l
600
S)<::l (c)
\e"""
.,-,; 'b~
~fl,
Axial fatigue Unnotched specimens Longitudinal and
400
50
transverse or-
200
ientations
o '--_-'--_...l-__....J...._----L_--''------''L-_-'--_-L.._--'--_----L_--'_ _-'----_...l-_-' -1200 -1000 -800
-600
-400
-200
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
:l
238
8-42. Types 600 and 329 Stainless Steel: S-N Curves for Two Processing Methods 1000 800 600
'"
o,
::2:
I I
.~
.L
400
I
I
I
IV
I
I I I. AISI 329 (electroslag remelted)
Vi
'" e
Ul
I
600 Steel (STAMP)-+--+--I
••
.
•
I
0.57
100 80
'\1'
60
0.51
40
'iii
-'" ",'
'" e
Ul
200 20 100 10'
10'
10'
10' Cycles to failure
Steel STAMP 600 Electroslag-remelted 329
Tensile strength, MPa(ksl)
Yieldstrength (0.2% olTset), MPa (ksi)
760 (110) 630 (91)
600 (87) 500 (73)
Mechonlcal properties Elongationin Reduction 50 mm (2 ln.), % in area, % 26
54
29
65
Impact energy,
strength,
J (R·lb)
MPa (ksl)
Fatigue
25 (18) 35 (25)
430 (62) 320 (46)
S-N curves showing test results and mechanical properties of STAMP-processed 600 stainless steel and electroslag-remelted AISI 329 stainless steel. Fatigue ratio (0 107/Rm) for 600 steel: 0.57. Fatigue ratio for electroslag-remelted 329 steel: 0.51.
Source: Metals Handbook, 9th Edition, Volume 7. Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, p 549
8-43. Grade 21-6-9 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rates AK, ksi • in.1/2
10
20
40
60
I
21-6-9
10- 4
Q)
u
> u ...... E E
22 to -196°C / (72 to -320°F) / -269°C (-452° F)
10- 3
2:
"t:l ...... III
/
"t:l
...oj.. III
s.. 0
Cl
.
u
III
>
10- 5
/ /
or.
~
Q)
U
10- 4
u
......u
,E 2:
"t:l ...... III "t:l
I
Q)
:l
Cl
'+:i III
u.
/ 10
20
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
239
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
Kromarc 58 Base metal Weld metal
......_ -::':::--_--:'-_....L.._----''--_ _.L------'
20
40
60
100
10- 6
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
10
20
30 40
> E E
200
10- 4
",-SMAW 24 and -269°C (75 and -452° F)
Gl
U
......u
60 80 100
10- 3
I
I
Z· 't:l ......
Gl
'"
u
't:l
> u ......
/"GTAW 24°C (75°F)
!l
'" ..r:
I
.~
~
I
0
~
'"
.~
10- 5 z· ~
'"
't:l
~
u
E u Gl
10- 4
: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
241
242
8-46. Duplex Stainless Steel KCR 171: Corrosion Fatigue
KCR 171 Whit. wat.r
400
pH ·4. IS
T • ISO·C 300 CI Q.
_ 0 .......
~
(\/
0 ...
200
l; ~
100
o
6 Hz
t:>
20 Hz 100 Hz
o
o
Nf
167Hz
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
9-1. Grades 200, 250, and 300 Maraging Steel: S-N Curves for Smooth and Notched Specimens 1500r - - - - - - r - - - - -.....----~----_,
CZZl Em
css '" ~
18Ni(300) 18Nj(250) 18Ni(200)
_ _-+-
1000
200
....=::j
150 'iii
c..
.:.l
v.
...e'"
If)
500
o
L.-
104
--'-
105
---L
106 Number of stress cycles
L.-
107
0 8 10 ~
Fatigue properties of maraging steels are comparable to those of other high-strength steels. Smooth-bar and notched-bar fatigue properties for I8Ni(200), I8Ni(250), and I8Ni(300) grades are summarized in the S-N curves shown above, 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. Improved fatigue properties can be obtained by shot peening and by nitriding,
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH. 1978, p 451
243
244
9·2. Grade 300 Maraging Steel: Fatigue Life in Terms of Total Strain
1
18% Ni morolling (300)
-1
10
,, \
--
/(J'f/E -2
10
--..
• - LOAD CONTROL
\ '\
- : TOTAL
~
~\
... A..-
-)-Trrr~ ..c
ELASTIC
\
PLASTIC~\ h
\
\
-3
10
L-J....l.l..JLlJJ.U-l.....LJ..J.JJJ.U.--I....LJ.J.LWL--I...1.LI..JWJJ..--L..LUJ..LJ.ll..-l-u..uJJ.lJ
1
10
Superposition of elastic and plastic curves gives fatigue life in terms of total strain. An actual example for this method of determining fatigue life is presented above for a maraging steel.
Source: Marc Andre Meyers and Krishan Kumar Chawla. "Mechanical Metallurgy: Principles and Applications," Prentice-Hall, Inc.. Englewood Cliffs NJ, 1984, p 700
10-1. Fatigue of Cast Irons as a Function of Structure-Sensitive Parameters
50
20
iii a:
§
10
•
...J
I&.:
• "B" BAR BAR
~ ·S~
. .1
I
10
MEAN FATIGUE LIMIT VERSUS (BHN) Fatigue of cast irons as a function of structure-sensitive parameters: Bhn, elastic modulus (Eo) and damping capacity (D).
Walter has shown that the fatigue properties of irons are highly dependent on volume of graphite and its morphology and distribution, as well as the matrix structure. He was able to reduce these factors to some easily measurable parameters, Eo, D, and Bhn, which gave good correlation with fatigue properties over a rather wide range of irons (see graph). It is reasonable that these parameters relate to fatigue performance, since they are measures offatigue-related properties. Eo, the modulus at very small strains, is controlled mostly by the volume of free graphite and to some degree by the graphite shape. Since the graphite present detracts from the matrix load-carrying area, the more graphite, the higher the stress on the remaining matrix-thus lower fatigue performance. D, the damping capacity, is controlled mostly by the graphite morphology and to some degree by the graphite volume. Sharp-edged flakes are greater stress raisers than rounded-edge flakes and spheroids; thus the higher the D, the poorer the fatigue performance. Bhn is largely a measure of the matrix hardness and, to some degree, ofthe graphite volume; thus the higher the Bhn, the better the fatigue performance. These easily measured properties are put to good use in industry as specification means and process-control criteria.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure. American Society for Metals, Metals Park OH, 1979, p 86
245
246
10-2. Gray Iron: Fatigue Life, and Fatigue Limit as a Function of Temperature Temperature, of
200
400
400
600
800
1000
I
I Fatigue life
Fatigue limit
350
-
- 50
-
- 40
-
-
300
250 IV
0-
:2
::;' 200
cil
150
, ". -,
e
<,
.
.,
'-
_____0
17
100
~
-
---- -% - " ~
-
<,~
50
"
20
1o
~
100
300
400
500
o
600
Temperature,oC
Number of cycles to failure
o Notched specimens
200
• Un notched specimens
• Notched specimens. stress based on net area
Composition: 2.84 C, 1.52 Si, 1.05 Mn, 0.07 P, 0.12 S, 0.31 Cr, 0.20 Ni, 0.37 Cu. (Ref 5) Typical fatigue life for as-cast gray iron of the above composition (left). Effect of temperature on fatigue limit for the same gray iron (right).
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 21
247
10-3. Gray Iron: S-N Curves for Unalloyed vs Alloyed
34
Plain Iron • Alloy Iron. 1% Ni. 0.4% Cr. 0.6% Mo. 1.0% Mn o
32
(f)
30
a. 0 0 0
..--
en
28 26
""",
'\.
230 220 210
0
~
200 190
•
180
.....
170
24
Knee
(f)
- /-----------1----0--
160
0
22
Endurance or ~ Fatigue Limit .,. - ._--_.1. ________________ •
20 18 10'
2
en (f)
(f)
L..
o,
10 5
-L
10 6
-
150 140 130
10 7
Cycles To Failure A fatigue crack starts in an area of high stress concentration after a large number of loading cycles. It is always a brittle type of fracture even when occurring in ductile metals. As the crack progresses it increases the stress concentration, and the rate of propagation under the cyclic loading increases. When the cr~ss section of the remaining metal becomes insufficient to support the maximum load, complete failure occurs as it would under an excessive steady stress. The number of stress applications that will induce a fatigue failure is less at higher maximum stress values, and conversely a larger number of stress cycles can occur at a lower maximum stress level before a fatigue crack is initiated. A plot of this relation for a metal is called an S- N curve and relates the maximum applied stress to the logarithm of the number of cycles for failure. When the number of cycles without failure exceeds ten million, the endurance life is considered infinite for bodycentered-cubic ferrous metals. The maximum stress that will allow this number of cycles is established as the endurance limit, or the fatigue strength or fatigue limit. Two typical S-N curves for a plain and alloy high-strength gray iron are presented above.
Source: Iron Castings Handbook, Charles F. Walton, Ed.. Iron Castings Society, Inc.. 1981, p 246
..... L..
(f)
248
10-4. Gray Iron: Effect of Environment
21 140
20 -
'ea.n
19
0 0 0
18
en
Fatigue Strength 18.000
130
co
Q.
PSI
~
1124 MPa)
120 17
en (JJ Q)
.... +-'
(JJ
Q) .... 16 +-'
110
(.f)
(.f)
15
_
14.500 DSt (100 MPal
100
14 10·
10·
10 7
10·
Number of Cycles The effect of various environments and corrosion inhibitors listed in the table below on the corrosion fatigue properties of gray iron.
Environment
Fatigue strength psi MPa
Air 17,920 Water 14,560 3% sodium chloride. . . . . . . . . . . . . . . . .. 5,600 1% borax 15,680 3% "Sobenite"* 17,920 3% sodium carbonate 17,920 3% soluble oil. 17,920 0.25% potassium chromate 17,920
Fatigue strength reduction factor
124 100
1.23
39 108 124
3.20 1.14 1.00
124 124
1.00 1.00
124
1.00
* "Sobenite" is a mixture oj 10paris sodium benzoate to 1 part sodium nitrite. The corrosion fatigue program involved testing in air, a spray of demineralized water, and a spray of three-percent sodium chloride solution; additional tests were made with a demineralized water spray and various known corrosion inhibitors. 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. Of the various alkaline inhibitors and soluble oils investigated, only borax was not completely effective for the pearlitic irons. Annealed ferritic gray irons were similarly affected by the sodium chloride solution.
Source: Iron Castings Handbook. Charles F. Walton. Ed., Iron Castings Society, Inc., 1981, p 255
10-5. Class 30 Gray Iron: Modified Goodman Diagram Mean Stress. MPa 50
100
150
200
30 ~---"""T---r--""---T""""'---"""'"2
200
Area of Finite Life 150 20
I--------+---~;<;;;;;;;;;;;;;;;;+.;;;;;;ijit----~
'(ij
a.
o o o
100
eo
CL
~
iii
Ul
....~ (f) 50
Cl
c
:il
0
S 0
-50 -101'-
o
.....L.
10
.l...-
20
...J
30
Mean Stress. 1000 psi
A modified Goodman diagram relates the endurance limit to an allowable working stress when it is superimposed on a steady stress. In many engineering applications, alternating stress is superimposed on a steady stress and requires special consideration. A method of relating the effect of the combined static and alternating stresses on the endurance limit has been developed into the Goodman diagram, of which a modified form is shown here.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981,P 251
249
250
10-6. Class 30 Gray Iron: Fatigue Crack Growth Rates Stress Intensity Factor Range. D.K in MPavm 10
20
40
60
100
Q)
o>-
Q)
o>-
~ o
~ E
s:
c
E
.~
c
10-'
z
Z
~
~
'"
"0
Band for Wrought Ferrite-Pearlite Steels
2iIII
a:
s:
~
Load Range
e
(9
i3
• 2000 o 2200 • 2500 o 2500 6 3000 ... 3300
10-8
'" U
Ib (910 kg) Ib (1000 kg) Ib (1130 kg) Ib (1130 kg) Ib (1360 kg) Ib (1500 kg)
'"
"0
10- 4 ~ III
a: s:
~
e
(9 .:>L
o
~
U
• 10
20
30 40
60 80100
Stress Intensity Factor Range. D.K in ksi VIiicli Fatigue crack growth rate. The endurance limit approach to design utilizes fatigue data taken on smooth, defect-free test specimens. For such specimens, 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. 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. 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,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. Only limited fatigue crack growth rate data are available on cast irons. These results are presented in the above chart for a class 30 gray iron, where dol dN is the crack growth per cycle and 6.K is the stress intensity range.
Source: Iron Castings Handbook, Charles F. Walton, Ed.. Iron Castings Society, Inc., 1981, p 250
10-7. Gray Irons: Torsional Fatigue for Various Tensile Strength Values MPa
1000 psi 60
400
58.000 psi (400 MPa)
/
50
en en
300 40
~
(j) "0 Q)
+-'
Q)
li} 200 a::
'0 .~ E
24.000 psi (166 MPa)
~
Q; 100 ~
o
14.000 psi (97 MPa)
..J
"0
c
1000 psi
Q;
~
0
::J
Mean Stress
-100
Torsional fatigue strength for three levels oftensile strength with various mean stresses.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 253
251
252
10-8. Gray Irons: Torsional Fatigue Data for Five Different Compositions 40
r---------------------., 1.50 en 1.25 Q)
E 30 E
.I: U
c
....OJ
.I:
1.0
.I:
0.75
-.J
....OJ
C Q)
-.J
C
20
Q)
~
U
~
lO
U
L..
lO
U
L..
o
U
0.5
lO
....0
10
lO
I-
0.25 I-
O'--__ '---""'--_-'----_-'-----I._-'-----'-----'400
300
----l_ _----l
800 1000
500 600
--'
2000
Number of Cycles Total length of six cracks (the first three cracks in each of two specimens of each iron) as a function ofthe number of thermal cycles between 1100 and 400 °C (590 and 200 ° C). Iron compositions are as follows: Composition. % Iron
A
B C
0 E
C
Si
Mn
Cr
Mo
3.43 3.45 3.45 3.44 3.43
1.65 1.74 1.68 1.69 1.66
0.57 0.59 0.63 0.58 0.58
0.49 0.30 0.21 0.50
0.30 0.38 0.39
Ni
Cu
Sn
0.60 0.97
0.59 0.87 0.30
0.077
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, pp 288, 289
10-9. Gray Irons: Thermal Fatigue-Effect of Aluminum Additions ,..----r----,---,----,-----r----r----.5
Gray Iron 3% AI. 0.6% Si (65% Ferrite) <> 2% Si .!: .150 o 0.5% Cu. 0.5% Mo D o c A 2% AI. 2% Si (4% Ferrite) .125 .175
D
+-'
.100 t---+----+---+---.,,>4F---t----+------l
Q)
....J ~
o
...ro
U
_ _--1
.075 t---+----+---.",..e..t----+-----.,.,.-=--
2
.050 I--------if------""--+---=-'f----+----+---'::=--'""=i
2
4
6
8
E
E
3
I
s:
g>
--+---t"7""'==------j4
10
12
Number of Thermal Cycles (X 1000) Thermal fatigue resistance of different alloyed gray irons.
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, demonstrating that a peariitic gray iron containing 3.4% carbon and 2% aluminum was highly resistant to thermal crack propagation.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 434
253
254
10-10. Gray Irons: Thermal Fatigue-Effect of Chromium and Molybdenum Additions 4
-
.15
1;..:-.,;.;;;, ..,. ,~~ ~.
.10
3.7% C 3
I
Bridge Cracked
.
2
/~
V
~ »-> Unalloyed ____~ CrNi _ '?"' . ~ .
..-'
.--
!
.-1--.05
I ...-:.
E
E
CrNiMo0'" CrMo(
en
.::.t.
~
~-.:
en
0
o ~
4
U
L.
U
'0
-
L.
---
3.2-3.3% C
s:
Q.
Q)
o
en Q) ..c o c
3
I Bridge Cracked
.15
..c
a. Q)
o
I
CrNiMo /
.
'0
.10
.
'
2
.05
o '--__-L._ _- - ' o 100 200
.L-
300
0 600
J..._ _----l_ _----J
400
500
Number of Cycles The depth of cracks resulting from the thermal cycling of gray irons between 860 of (460°C) and room temperature.
Alloying with molybdenum and chromium provided superior thermal fatigue resistance compared with irons that contained other alloying additions or no alloying at all. In this case, the improved thermal fatigue resistance is believed to be directly related to the higher elevated-temperature tensile strength and better stability of the chromiummolybdenum irons. However, it must be remembered that this improvement is related to and dependent on the temperature cycle and base iron composition, as shown above. It has also been indicated that the development of an acicular matrix structure, by adding relatively large quantities of molybdenum and copper, supplies a less than desirable influence on thermal fatigue cracking.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Casings Society, Inc., 1981, p 288
10-11. Gray Irons: Thermal Fatigue-Room Temperature and 540°C (1000 OF) Tensile Strength. MPa
o
100
200
300
1500
Q)
0 >-
o
'0
1000
Q; .D
E :J
Z
500
o Lo
' - -_ _---'
10
20
--'-
-'-
30
40
'---'
50
Tensile Strength. 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 .
For good resistance to thermal fatigue, gray irons should have high thermal conductivity, a low modulus of elasticity, high strength at both room and elevated temperatures, and for temperatures above 900 OF(500 "C), resistance to oxidation and structural change. Because some of these properties are in opposition, a compromise must be made in selecting the most appropriate metal for each type of service. As the maximum temperature to which the gray iron is cycled and number of cycles increase, the number and size of thermal fatigue cracks become larger. The above curves illustrate the influence of room- and elevated-temperature strength on the thermal fatigue resistance of irons having similar carbon equivalents, thermal conductivities, matrix structures, and elastic moduli. Those irons with higher room and elevated-temperature tensile strengths (achieved by alloying) generally display higher thermal fatigue strength.
Source: Iron Castings Handbook. Charles F. Walton, Ed .• Iron Castings Society, Inc.• 1981, p 285
255
256
10-12. Gray Irons: Thermal Fatigue Properties-Comparisons With Ductile Cast Iron and Carbon Steel Maximum Cycle Temperature. F 1200
1400
1600
1o',---,--------,r--------,---,
z d> c
:;;; 10'
Steel
o
~
(J
Ductile Iron
B
u
>-
(J
'0
Gray
AI ; 1m",
Q> 10'
.0
E :J
Z
16~0.,...0--------=-.L------~-----~~ 700 800 900
Maximum Cycle Temperature. C
The above curves show the variation of the number of cycles to cracking with the maximum temperature of the cycles for gray iron, ductile iron, and carbon steel. Compositions of the four gray irons are as follows: Analysis. % Iron A B
C D
C
Si
Mn
Cr
Mo
Other
3.43 3.49 3.48 3.50
2.37 2.37 0.60 2.38
0.78 0.84 0.88 0.83
0.22 0.24 0.23 0.30
0.32 0.22 0.20 0.77
0.21 Sn 2.37 AI 1.51 Cu
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, pp 286, 287
10-13. Cast Irons: Thermal Fatigue Properties for Six Grades
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. Compositions are tabulated below. %Mg
Analysis
%C
%5
%Mn
%P
Class 35 Gray Iron
2.96
2.90
0.78
0.07
Ferritic Compacted Graphite
3.52
2.61
0.25
0.05
0.015
Pearlitic Compacted Graphite Ferritic Ductile
3.52
2.25
0.40
0.05
0.015
3.67 3.60
2.55 2.34 4.84
0.13 0.50
0.06
Pearlitic Ductile Alloyed Ferritic Ductile
0.030 0.030
3.48
0.31
0.05 0.07
Alloys 0.12Cr
0.030
1.47Cu
0.54 Cu 1.02Mo
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society. Inc., 1981, pp 393, 396
257
258
10-14. Ductile Iron: Effect of Microstructure on Endurance Ratio-Tensile Strength Relationship Tensile Strength. MPa 300
~o
J.
500
700
900 I
I
1100
I
1300 I
0.5 -o\'ox>
~.~ "\~t x Pearlitic
Ferritic.~.
o .."
.o~ "x~x: x o ,
'"
II: Q)
g
0000
0.4
~
::J
'U
c
~ \ ·K
Tempered Martensite
<,
W
• j( .....
0.3
• • ••
60
100
140
160
Tensile Strength. 1000 psi
In general the fatigue limit for ductile iron increases with tensile strength, but as with other ferrous metals, the increase is less than proportional. The relation between the tensile strength and the endurance ratio for the annealed, ferritic irons is different from that of the irons with a matrix of pearlite or tempered martensite, as illustrated above.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 341
10-15. Ductile Iron: Effect of Microstructure on Endurance Ratio-Tensile Strength Relationship Tensile strength, ksi
50
75
100
125
150
175
0.51-----e-..... "ki~---T___+-----__1I_-----+__-----__+--_t
..o, ~
1ie
O.4I------_+_---'''--------'''r__+-----.-+--'''o,~
......I _ - - - - - _ + _ - - - - - _ _ + - - _ _ 1
::>
"0
<:
w
0.31------t------+-----t--------1r---'''''''=----+----j
• 200 Tensile strength, MP.
The influence of tensile strength and structure on the endurance ratio of ductile iron is indicated in this graph. Endurance ratio is defined as endurance limit divided by tensile strength. Because the endurance ratio of ductile iron decreases as tensile strength increases, regardless of structure, there may be little value in specifying a higher-strength ductile iron for a structure that is prone to fatigue failure. For tempered martensite ductile iron, the improvement in fatigue strength due to an increase in tensile strength is greater than for pearlitic or ferritic structures. This is indicated in the graph above by the shallower slope for martensite.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels. American Society for Metals, Metals Park OH, 1978, P 45
259
260
10-16. Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades, Using V-Notched Specimens 400'.---------,-------,-------,---------,
3S01-------1-------+------+---------1 F_I6G-40-18 .......odl
~v...otchod
3ool----------+------+-------t--------I
g.
~
2001--------=zilJ!l ISOI--------+----==
20 loo'f--------f-------+-------1-----------l
106
Fatiguelift. cyeles
400,-------.-----------,--------,r--------, 3S01-----------lI----------l--------l'---------l Pllrlitlc (80-66-06 •• -e...t) 46° V-notched
3001-----------1r--------l--------l---------l 40
.r.
~
~
-6 2SOI -- - - - - -+----:
~e
i!'
~o
:;
1
il. 2001-----------1r----"
:
ISOI---------1I--------f--------f---------j
20
1001-----------1r-------f-------1--------l
106 Fatiguelife. cycles
Top:S-N curves, including scatter bands, for annealed ductile iron. Bottom: Similar to above except for as-cast pearlitic ductile iron. All test specimens were V-notched (45°).
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 43
10-17. Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades, Using Unnotched Specimens .00r-------,.-------,--------,.-----------, F.,rillc: (60.4o-18 ann••IId) unnotc'*l
3501--------t------+------+--------I 3001--------t----::;
e:lE
'0 ]
~
2501--------t--
S
~
~
= ~
J
= 1
:
200
1501--------t------+------+--------I
20
1001--------+------+------1-------; 106
Faligue life. cYcles
.OOr--------,------.----------,r--------, 3501---------f~
3001--------;------' t:.
'0
:lE
ii 2501--------;-------+------......,1--------1
~
go
a ~
J200
1501-------+------+--------1------.,
20
1001-------+------+--------1------., 5:0'"'·-------'<--------';-------:--',--------' 106 10' Faltgue lIle,cycles
Grade
Tenolle alrenl!h MPa kal
Unnolched Endurance UmII Endurance MPa kat rallo
60·40-18 80-55-06
480 680
205 275
70 99
30 40
0.43 0.40
Notched Endurance Umil MPa kat
125 165
18 24
Endul' ance ratio
Slreaa coneenlrallon lacier
0.26 0.24
1.67 1.67
Top: Similar to upper graph on the opposite page, but here the specimens were unnotched. Bottom: Similar to lower graph on the opposite page, but here the specimens were unnotched. Data in table pertain to graphs on this and the opposite page.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, p 44
261
262
10-18. Ductile Iron: Fatigue Diagrams for Bending Stresses and Tension-Compression Stresses MPa 1000 psi
MPa 1000 psi
100 600
600 80
400
·E
60
JOO
40
~
::; ~ 200 c:
5
5
400
200
Q)
o
c
5 100
20
-0 c:
-0
c
W
W
o -20
-100
-40
-200
~--:-f\P'A"':ld'-?-:!:---!c:---f::---,~~I:---
1001 psi
-200
200
400
600
-JO -40 L-_-:-'-:-_-=-~_--:-'-,----_-,-:-_-=-'-MP, 100 200 JOO 400 600 o
MPa
Mean Stress
Mean Stress
Fatigue diagrams indicating endurance limits for five grades of ductile iron under bending stresses (left) and tension-compression stresses (right). Minimum properties of the irons are given in the table below.
Iron No. 1 2 3 4 5
Min. Tensile Strength
Min. Yield Strength
1000 psi
MPa
1000 psi
MPa
Min. Elongation Percent
55 61 72 87 102
38 42 50 60 70
36 41 51 61 72
25 28 35 42 50
17 12 7 2 2
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, pp 344, 345 and 346
263
10-19. Ductile Iron: Effect of Surface ConditionsAs-Cast vs Polished Surface Tensile strength. ksi
90 100 120 _':"';'::'_...:...r:.-....,....;.,~-, 50 350 n60r--_":,,,_-.--~......,.._--, .........._r----r-.......... g
n8
'" 3001---+----+---i--7"'q...---f---+-----i
a.. ::2:
45
40
.<:='
200 1----+---+--7"'~_+_---' 25 150
_ __'___ _-'-_ _ 400 600 500 700 ~
__L. _ _- - - - '
800
" ' - _ ___'__ _ __ '
900
1000
1100
Tensile strength, MPa
Tests made on 10.6-mm (O.417-in.) diameter specimens. Fully reversed stress (R = -1).
Data given in the above graph show that the endurance limit for any given strength level of ductile iron is significantly affected by surface conditions of unnotched specimens. The endurance limit is much higher for the polished specimens than it is for the as-cast specimens, which have relatively rough surfaces.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, p 45
264
10-20. Ductile Iron: Fatigue Limit in Rotary Bending as Related to Hardness iii
60
0-
...
... •
'E ::i
,.,
0)
';:; 40
"0 Q)
lD
30
e
0
a:
~
360
••
20 100
----I-
...
'E
::i Q)
300
::J 0)
';:;
• • • • •
.,....•
0)
c
0...
~
Q)
::J
c
------
••
. .. ..... •• •
60
400
•
0 0 0
260
0)
.s
"0
c
Q)
• • • •
lD 200
~
s0
a: 160
200
300
400
Hardness. Brinell 'iii
00 0 0
60
0...
- 400
60 -
Q)
::J
/
0)
.~
u, 0)
c
40
r-
"0
c
Q)
lD
z-
0
a:
30 100
.
/
...
'E ::i
~
.
A. I
200
.~
..
~
...
'E
.,-
::i 360
Q)
::J 0)
';:;
u,
300
0)
.s "0
c
Q)
260 lD
z-
I
I
I
I
300
400
600
600
0
700
a:
Micro-Vickers Hardness Number
Top: Relation between Brinell hardness and fatigue limit in rotary bending for ductile iron. Bottom: Relation between rotary bending fatigue limit and matrix hardness for ductile iron.
Source: Iron Castings Handbook, Charles F. Walton. Ed" Iron Castings Society, Inc" 1981, p 347
10-21. Ductile Iron: Effect of Rolling on Fatigue Characteristics
Rolling Pressures (Pounds)
75
70
500
406 580 768
65
60 400
'(ii
o,
o 0 0
en (J)
~
,,
55
50
<0
n, ~
,, "
"", "
u:;
83 "
45
"-----------j
,----------- - Unnotched. Unrolled
300
40
35
30
10'
Unrolled
10'
10'
200
10'
Number of Cycles of Stress (Log Scale)
Fatigue strength of ductile iron can be increased substantially by cold working, especially when this method is applied to stressed radii or notches. 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. The improvement in fatigue properties obtained by various rolling pressures on ductile iron is indicated above.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc .• 1981, p 348
265
266
10-22. Ductile Iron: Effect of-Notches on a 65,800-psiTensile-Strength Grade
I 36
<, 32
'iij 0.
0
I
1
t 0.750"
0
<,
2.75"R
I
- 250
i
0}17" 10.4'72"
~
t
..
I
-
225
-
200
I
0"",
~ ~()OOO
Unnotched
-"'0-
28
o o o
eo
n,
0 '\.
'\.0"
~ 20
W
t
I'"
\
W (J) ~
Ul
~
-i
0 o
175
"1
!
~
-
0.417"
0.700"
V-Notched
~
-
150
.....
0.25 mm Root Had.
0
125 1--0-000--
16
10'
10'
10'
100
10'
Number of Cycles
The unnotched and notched fatigue properties of an annealed ductile iron with a tensile strength of 65,800 psi (454 MPa). The endurance ratio is 0.41 and the notch sensitivity ratio is 1.67.
Source: Iron Castings Handbook. Charles F. Walton, Ed.. Iron Castings Society, Inc.• 1981, p 341
10-23. Ductile Iron: Fatigue Crack Growth Rate Compared With That of Steel Stress Intensity Factor Range. t.K in MPa
v'rri
20
Q)
Q)
<3
<3
>-
>-
o
o
<,
:? 10-'
E E
o c c
Band for Wrought Ferrite-Pearlite Steels
Z "0
<,
'"
c
10-'
z ~ eo "0
"0
2
'"
a:
s:
s:
~
~
e
o (910-'
• 3300 Ib (1500 kg) o 1650 Ib (750 kg) .. 2700 Ib (1225 kg) • 2200 Ib (1000 kg)
.:.!
o ~ U
<.9 .:.!
o
~
U
10-' '--_...J......_ _...L-_L-...L-..I...-I....L...I..J....L-_ _..I...-_L-.L.....I 10
30 40
60 80100
Stress Intensity Factor Range. t.K in ksi yinch
Fatigue crack growth rate of annealed ferritic ductile iron, compared with that of ferritic-pearlitic steels.
Source: Iron Castings Handbook. Charles F. Walton. Ed.. Iron Castings Society. Inc.. 1981. p 349
267
268
10-24. Malleable Iron: S-N Curve Comparisons of Four Grades 60
r--------...,---------, 400
350
50 .iii 0.
c...
0 0 0
:2: -0
300 ~
-0
--l
0 --l
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, low-cycle fatigue. The results with a 95% confidence limit are presented in this S-N diagram. Unmachined and notched surfaces do reduce the fatigue strength. The reduction factor is as low as 1.2 for the lowerstrength irons to over 2.0 for the higher-strength irons. Inducing compressive stresses into the surface by rolling, coining, or shot peening can increase the fatigue life of a component significantly. Design with adequate sections that are well blended to reduce stress concentrations is most effective in reducing the possibility of a fatigue failure.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981,p 311
10-25. Pearlitic Malleable Iron: Effect of Surface Conditions on S-N Curves 70 '00 60
400
C-
o 0 0
<0
a.. ~
60 300
enIII
40
Vi
30
en
III
~
Vi
~
200
20 10'
10'
10'
10'
Number of Cycles The influence of as-cast surfaces, smooth machined surfaces, and machined notches on the fatigue behavior of pearlitic malleable irons. Iron 1 is grade 60003 and Iron 2 is grade 80002.
Surface finish has an important influence on fatigue properties, as shown above. Samples of malleable grades 60003 and 80002 were tested in fatigue with "as-cast" and machined surfaces. Samples of the 60003 grade were also included with a machined surface containing a sixty-degree notch that was 0.050 in. (1.25 mm) deep. The resulting data are shown in this diagram.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 310
269
270
10-26. Pearlitic Malleable Iron: Effect of Nitriding
1 PIECE
t----50.0
9 PIECP.S
ATHOSPIlr.RE
NITRIDED
COHPRESSIOH 11,160 POUND~ TENSION
7.600 POUNDS
1,01.--------------------------------II) 5
HUMBER OF CYCLES
Effect of gaseous atmosphere nitriding on fatigue characteristics of pearlitic malleable iron, tested by tension-compression.
Samples were austenitized, oil quenched, and tempered to 241-269 HB prior to nitriding or testing without nitriding. This chart indicates an increase in fatigue life of 750,000 to 2,700,000 cycles attained by nitriding.
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals. Metals Park OH, 1977, P 287
10-27. Ferritic Malleable Iron: Effect of Notch Radius and Depth Depth of notch, in.
250
o
002
004
006
008
~ tNotch
rad.ius o 0.13 mm or 0.005 in.
~
c,
<,
:;;
-5 0>
150
.~
I
I
I
100
~ . r--
u.
~
50
o
o
30
~
• 0.75 mm or 0.030 in.
...............
~
t;
'" 5,
I
0.25 mm or 0.010 in.
200
--
-
E: 20
~
'" 5,
•
.~
r--- r-s-
u.
10
o 0.5
1.0
1.5
2.0
Depth of notch, mm
Effect of notch radius and notch depth on fatigue strength of ferritic malleable iron.
Fatigue strength of unnotched ferritic malleable iron is approximately 50% of the tensile strength, or from 170 to 205 MPa (25 to 30 ksi). The graph above summarizes the effects of notches on fatigue strength. As a rule, notch radius has little effect on fatigue strength, but fatigue strength decreases as notch depth increases.
Source: Metals Handbook. 9th Edition. Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 65
271
272
11-1. A286: Effect of Environment
A286 w
C)
z : z c
VACUUM
o AIR 593·C • VACUUM 593·C AVACUUM 20·C
0.01
a:
~
en
~ 0.001
en
c
-J
Q.
Plastic-strain range versus fatigue life for A286 ferrous alloy in air and in vacuum af 593°C (1095 OF). Numbers adjacent to test points indicate frequency in cycles per minute. Note absence of frequency effects in vacuum.
Coffin has suggested that for a number of materials, virtually aU of the degradation in fatigue life at elevated temperatures can be attributed to environmental interactions. He noted that frequency effects in the low-cycle-fatigue law could be eliminated for a large number of metals and alloys by testing in vacuum (note above). Additionally, it was noted that tests performed in vacuum showed transgranular crack nucleation and propagation versus intergranular nucleation and propagation in air at elevated temperatures. These results are not unambiguous, since Koburger has shown a frequency effect in high-cycle fatigue for directionally solidified eutectic alloys when tested in air and in vacuum, particularly at elevated temperatures. The primary difference in these results may be related to the lack of intergranular cracking in eutectic alloys.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 343
11-2. A286: Effect of Frequency on Life at 593°C (1095 OF)
/.--------- - ------
~
~---
l!>;:;::--:: /' --1>-;;:::::::0 /' /'
:0
-:
".0
,/
/
) /
A 286-593°C Kr=3.0 .MI. =60ksi 2
--:;....--:;;.--
--
v STANDARD HT-AIR oHT#1 -AIR • HT#I -VACUUM 0 HT#2 -AIR I> HT #3 -AIR 0 DS HT # 3 -AIR ~DS-STD HT - AIR
Effect of frequency on life of notched fatigue bars of A286 at 593°C (1095 OF) in air and vacuum. As indicated, decreasing frequency has a degrading effect on fatigue life ofsamples tested in air, with little or no effect on samples tested in a vacuum.
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 13
273
274
11-3. A286: Fatigue Crack Growth Rates at Room and Elevated Temperatures AK, ksi • In.1I2 10
20
30
40
60
SO 100
t
A' /
10-3
53S·C
(1000.F)~'
427·C (SOO·F) ,
,/
/
/
(i / I"
10-4
,
A
/
/
10- 4
/
/
01
/ "'-24·C (75·F)
10-6
U
~
.5 2'
~ "tl 316·C (600·F)
I, / ill
10-6
, I
10-6
I
A·2S6
I 10
20
30
40
60
SO
100
Stress-intensity factor range, AK, 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 0.67 Hz (elevated temperatures), anR ratio ofO.OS, and at L-T, T-L, R-L, and R-C orientations.
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, which precipitate as intermetallic compounds such as Ni, (AI, Ti) and Ni 4Mo(Fe, Cr) Ti on aging. Various mill forms of the alloy are usually supplied in the annealed conditionCondition A (980°C, or 1800 OF, for one hour followed by quenching in oil or water). Precipitation hardening occurs on aging in the range from 700 to 760°C (1300 to 1400 OF)for 16 hours. Other combinations of heat treatments may be used depending on the application. One variation is to re-solution treat at 900°C (1650 OF)for two hours, quench in oil or water, and age at 700°C (1300 OF) for 16 hours. This variation results in improved room temperature properties but less desirable stressrupture properties.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, p 161
11-4. Astroloy: S-N Curves for Powder vs Conventional Forgings 100 Conventional Forgings
'"I 2
.
o
I
Powder Forging
80
x
'
~
\I) \I)
w
60
...
'" >
...
0
40
s
s
20
0 104
10 5
10 6 CYCLES
S-N curves for conventional and powder forgings of Astroloy (notched versus smooth),
Testing was performed using standard methods at 705°C (1300 OF) and a combination of steady and vibratory stresses for which comparative data were available. Cycles to first indication (crack) were comparable to conventional material. Crack propagation as judged by the number of additional cycles from first indication to failure was slower than conventional material, as shown above.
Source: M. M. Allen, R. L. Athey and J. B. Moore, "Application of Powder Metallurgy to Superalloy Forgings," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 97
275
276
11-5. Astroloy: Powder vs Conventional Forgings Tested at 705°C (1300 OF) Steady Stress = Vibratory Stress = 40,000 psi
x =
Crack
o = Failure Powder Forging
Conventional Forging
10 4
CYCLES
Astroloy tested in high-cycle fatigue at 705 °C (1300 OF). Vibratory stress levels were selected to facilitate a direct comparison between conventional and powder forgings.
Source: M. M. Allen, R. L. Athey and J. B. Moore, "Application of Powder Metallurgy to Superalloy Forgings," in Source Book on Powder Metallurgy, Samuel Bradbury. Ed., American Society for Metals, Metals Park OH, 1979,p 97
277
11-6. FSX-430: Effect of Grain Size on Cycles to Cracking lSI 4)0
large grains
0.4
, ~ ,
\
~
if
\
•
'"Z
~-LG
I I I
HI MIN
\
'\
C
II:
t---O-I- IG
fine grain a
.
~
." 0.)
c
N
0.01 10.1~",1
HZ
AVG MAl
..
i
\
\
II:
~ ~cb
0.01
(O.~'I
....& ...... ~
(
\
~,
.,; ~ 0 c
\
II:
E E
II:
>
0.05 10.7621
... c ... II:
~
!:
~
0.0"' 11.0161
"-
...
II:
L
0.2
o
40
.0
IZO
160
ZOO
240
zeo
no
CYCLES TO CRACK
SoN curves for alloy FSX-430, showing effect of grain size on cycles to cracking,
Source: Eric Bachelet and Gerard Lesouit, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed.. American Society for Metals, Metals Park OH, 1984, P 336
278
11-7. FSX-430: Effect of Grain Size on Fatigue Crack Propagation Rate ,",,,,/CYCLE I O.OJ
GROWTH RATE IN llNlAl! flAIlGE ODZ
o
0.01
0.04
~
J--O-1-fG 0.4
~-LG
I
I
I
I
CRACK AVG CflACK
I
0.01
t-{j----I
2
. '"
I
~ z·
:ca:
I
...
I
a:
I
<
0.3
....
0)1
,
<
C
...%
..
!: c
.
'" C
0.03 II: (O.TUI~
)
e,
II:
)~
<
Q,
,,
It'
IE
...i
O.O!
(o.~oeio
I
~
_
(O.Z:>41 E E
I
0.04
~
(1.0151
I
large grlolinB
0.2
o
I fine grains
0.5 1.0 1.5 GROWlH RAlE IN LINEAR flANGl (IO·"n.lCTCLEI
2.0
Fatigue crack propagation rate-effect of grain size on fatigue characteristics of FSX-430.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of SuperalIoys," in Superalloys: Source Book, Matthew J. Donachie, Jr.. Ed.. American Society for Metals. Metals Park OH, 1984, P 337
11-8. HS-31: Effect of Testing Temperature 70
!
60 .u; a.
50
..........
o o Q 40 30 20 1100
Aged 50 hr at 1350 F
(
""""-<
Fatigue strength 100 million cycles
"'- r---.....
(["---...:
--
:--
1400 1200 1300 Testing temperature, F
1500
Effect of testing temperature on fatigue strength of HS-31 casting alloy, after aging at 730°C (1350 OF), for 100 million cycles.
Source: ASM Committee on Heat-Resistant Castings, "Heat-Resistant Alloy Castings,"in Source Book on Materials for ElevatedTemperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 237
279
280
11-9. IN 738 lC Casting Alloy: Standard vs HIP'd Material
y-"STANDARD CONDITlON" MATERIAL
220
<,
CII
200
I
E E
:z;
.
180
Po.
+1
Po. ......, Ul Ul
160 11M>
•
'-.<,
<, • <,
MATERIAL
<, -"<.
"
...... ......
~
E-i Ul
<,
..............
120
-
tI$..............
107 CYCLES TO FAILURE SoN curves for casting alloy IN 738 LC. High-cycle fatigue properties ofnimocast alloy IN 738 LC tested at 850°C (1560 oF).
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Super alloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, P 340
11-10. IN 738 LC: Effect of Grain Size on Cycles to Failure
240 /
220
N
I
E E
.
2:
200
a..
+
I
"
STANDARD CDNDITIDN
-,
•
.
......
<,
.......
•
FINE GRAINS
.......
180
a.. 111 111 W 0:
t-
111
160
•
......
.......
......
• .......
•
140
MATERIAL
......
'
• ....... ~
.
120
•
-
•
e· 10
6
10
7
10
8
CYCLES TO FAILURE SoN curves for alloy IN 738 LC. High-cycle fatigue properties of extra-fine-grain and conventional material tested at 850°C (1560 OF).
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book. Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, p 340
281
282
11-11. IN 738 LC: Effect of Grain Size on Cycles to Cracking 'N - 738
0.4
.......-o--i ~
I I
M'N. HZ N ••,
fG
-LG
0.0' o.ZS41
I
MAX."2
'i
!
.!i
...
~
a:
•
w
0.02
_ O.J
IO.!lOI1
!
... :... f .... c e::
, ~~)-~......--
~
c
-
....
o
eo
.20
160 200 CYCLES TO CRACK
oJ
:tIO
... ...a: 2:
~
0.04
11.0161
fine grains
2.40
::l
c a:
c 0.03 a: 10.JUI
~
large grains
0.2
.; 0
320
S-N curves for alloy IN 738 Le, showing the effect of grain size on number of cycles to cracking.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, P 335
L
283
11-12. IN 738 LC: Effect of Grain Size on Crack Propagation Rate c.ROWTH IlATE IN L1I1EAR 11&",[ l ..." u n l l
O.GI ,
0.01 ,
O.OJ I
0,04 ,
•. 4 0.01
O.~~I
E
!
s
t;
.:
O.GZ
(0.!>08)
.0.) f-
. ..•.... ... !
&
..•~
Ci
~
QOJ
(0.7611
.,
f-
........
..'" %
0.0-
f
•
1.01'1
...
.....
i
IN ' ) I
L
t-O-i :E-C»i -
0.2 f-
I
,
I
I
o
0.'
1.0
I
I
FG
LG
CRACI( I /IIIG CU~K I
1.5
1.0
GIIOWTH RAT( IN LINEAR RANG[ llli'in.lCYCLL,
SoN curves for alloy IN 738 Le, showing the effect of grain size on crack propagation rate.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OB, 1984, P 336
284
11-13. IN 738 LC: Fatigue Crack Growth Rate at 850°C (1560 OF)
10- 5 do dN (m/cycle) 10- 6
100Hz
• Alloy I IN 738 LC Temperature I 850°C .0
10- 10
Abo coarse - grained
R
10- 11
fine-grained
t'
0.1
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.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book,' Matthew J. achie, Jr., Ed., American Society for Metals, Metals Park OH, 1984,p 341
000-
11-14. Inconel 550: Axial Tensile Fatigue Properties in Air and Vacuum at 1090 K 50 40 30
....
C>
2
(a)
20
'iii 0-
vi VI w
a:
10 50
t-
·VACUUM
VI
Z
w
°AIR
40
::E
(b)
30 20 10 105
10
107 LIFE. cycles I I II lid I I "" II 100 1000
10000
LIFE. h
Axial tensile fatigue properties oflnconel550 at 1090 K in air and vacuum. (a) Ratio of cyclicto mean stress= 0.125. (b) Ratio of cyclic to mean stress = 0.667. Testing frequency = 33 Hz.
In reversed bend tests on lead at 500 cycles/min, Snowden demonstrated a difference oftwo orders of magnitude in fatigue life between vacuum, air, and pure oxygen. At all strain levels vacuum endurances exceeded those in air, which exceeded those in oxygen. Intermittent stress-free exposure to air had no effect on the lifetime in vacuum. At high temperature (l090 K) vacuum also improved endurance, relative to air, of the Co-base alloy S-816 and the Ni-base alloy Inconel550, although the effect was much smaller than that seen in lead. Endurances for the nickel-base alloy converged at low stresses, indicating a possible strengthening effect of air, as shown above.
Source: R. H. Cook and R. P. Skelton. "Environment-Dependence of the Mechanical Properties of Metals at High Temperature," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 81
285
286
11-15. Inconel 625: Effect of Temperature on Cycles to Failure 600
500 427°C
Joo
80 of)
,~ e:--
400 <0
a.
:;;
Ii 300
~
200
'-
-
<, ...
---
29°C (85 of) 538°C (1000 °F)_ 60 649°C (1200 of) I-
NOICh~~ specimens (Kt~3.3)
760°C ('1400 of)
i
29 Oc (85 of)
~
871°C (1600 of) _
20
100
Cycles 10 failure
S-N curves for hot rolled solution treated Inconel625 bar 15.9 mm (0.625 in.) in diameter at various temperatures. Average grain size was 0.10 mm (0.004 ln.),
Source: Metals Handbook, 9th Edition. Volume 3. Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 143
11-16. Inconel 706: Effect of Temperature on Fatigue Crack Growth Rate Stress-intensity factor range, aK, ksi • in.1/2 10
20
30 40506080 100
200 300
10- 4
-196°C (-320°F) Ql
10-3
U
Ql
> u
U
--EE z --
~
"C
"C
III
III
"C
10- 5
l!l III
..
Ql'
s:
i e
i e
-269°C (--452°F)
Cl
u
"C III
s:
~
--z·,E -.... Cl
10- 4
~
u
I! u
I! U
Ql
Ql
::J
Cl
::J
III
'':;
Cl
'':;
III
II..
II..
10-6
20
30 40 50 60 80 100
200
300
Stress-intensity factor range, aK, MPa . m1/2 Fatigue crack growth rates of Inconel 706 forged billet (vacuum induction melted/vacuum are remelted) at an R ratio of 0.1 and a frequency of 10 Hz. Heat treatment: 980°C (1800 OF) I h, AC; double aged 730°C (1350 OF) 8 h, FC to 620°C (1150 OF), hold 8 h, AC.
Results of FCP tests at room temperature and at temperatures as low as -269°C (-452 OF) for Inconel 706 are shown above, At equivalent 11K values, the fatigue crack growth rates for this alloy are slightly lower at subzero temperatures than at room temperature,
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties ofSuperalloys,"in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, p 297
287
288
11-17. Inconel"713C": Effect of Elevated Temperatures on Fatigue Characteristics
"in
50
c.
g 4 a I-+-+-f-Ht--~H-Yt---+--+ o
:i 3 a 1--+-+++t--~,*h-r--+--P't"+l:>---1--+t-t-l Q)
"-
iii
Ol
2 a I--+--+--r-f-..---r---t-+-+--i---i--t-++t---+--t-+-+i
~ o
E 10 .2?
NOTE: Higher fatigue strength at 1500 F than at 1200 F is consistent with tensile strength relations in graph shown above left.
OL-..I...-...L....J.....LJ...---l...~ .............1...--'--...L....J.....LJ...---L.----L~
0.1
I 10 100 Millions of cycles to failure
SoN curves for Inconel "713C." Tests were performed at two different elevated temperatures as shown.
Source: ASM Committee on Heat-Resistant Castings, "Heat-Resistant Alloy Castings," in Source Book on Materials for ElevatedTemperature Applications. Elihu F. Bradley, Ed.. American Society for Metals. Metals Park OH. 1979, P 235
289
11-18. Inconel "713C" and As-Cast HS-31: Comparison of Two Alloys for Number of Cycles in Thermal Fatigue to Initiate Cracks
Thermal fatigue Cycles to first crack
Material
HS-31 Inconel
ri
l-rnin cycles, 100 to 1700F
Material avg1
avg
-1&
1713C"
a
,r I Thermal fatigue Cycles to develop . 8I-m, erne k
HS-31 Inconel
I
2
Thousands of cycles
3
IIIII
1713C"
3 tests each motertol
a
I
2
Thousands of cycles
Thermal fatigue properties of HS-31 compared with those of Inconel "713C." Left: Number of cycles required to initiate cracks. Right: Number of cycles required to develop VB-in. crack.
Source: ASM Committee on Heat-Resistant Castings. "Heat-Resistant Alloy Castings." in Source Book on Materials for ElevatedTemperature Applications. Elihu F. Bradley, Ed.. American Society for Metals. Metals Park OH. 1979, P 235
3
290
11-19. Inconel718: Effect of Frequency on Fatigue Crack Propagation Rate I
5 X 10-3
I
I
I
I
I
-
f-
o 2
o 5 X 10- Hz } 05 X 10-' H,
0
o",~/ / •
\
I
o
5 X 10-4 -
1>
/S /ili.
o 00
-
&
/
o
/;".
o
/&
o
o 5 X
10-5 I-
tL {--
a~a / I.•
o
I /
a
&
a
~/
20 Hz 2Hz
_
0.5 Hz
&&/ &
//
& ......---/
5 X 10-6
/ &1 1..----I._ _--1.._--'_...L...--1..--'---'--'-...L........L.---l
8
10
15
20
30
40
50 60
Stress-intensity factor range, 61<, MPa . 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.
The effect of frequency at 550°C (1025 OF) was studied using a sine wave; the results are shown above. Below 0.5 Hz, the FCP rate was more rapid, and the crack surfaces showed an increased amount of intergranular fracture with decreasing frequency, with the crack path following the boundaries of the largest grains. One may be inclined to attribute the increase in FCP to either creep or environment, but this may not be the case, because different modes of deformation may have occurred at different strain rates.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 294
11-20. Inconel718: Relationship of Fatigue Crack Propagation Rate With Stress Intensity
• 25°C (77°F) • 550°C (1025°F) 20 Hz
• • ••
5 X 10-4
• • •• • • .. •• ••• • • • • • • • ••
.
5 X 10-5
.. ,
•• ••••• ••• •
:-• •
5 X 10-6
••
•
._~
: Twins
10- 6
......--L
8
10
No twins
....L._--'-_.J...-............L.....L.........L.-............
15
20
30
40
50 60
Stress-intensity factor range, Li K, MPa . 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.
An important effect is the hydrostatic state of stress in the tip region. This idea has been considered for Inconel 718. The FCP response is shown above for 20 Hz, where the effect of temperature is to increase the FCP rate, especially at LiK levels.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 290
291
292
11-21. Inconel 718: Relationship of Fatigue Crack Growth Rate With Load/Time Waveforms
a K. ksi • in. '/2 10
20
30
40
60
• /'v (2) 5 X 10-2 Hz
{
2 Hz
_ rxr
(1)
/'v
(3)
D
I
(2)
I
rf·
Sinusoidal ~ ~
2 5 X 10- Hz 5 X
10- 4
I (1)
'Jr¥
~ - / (3)
-;0
2X
I.• _ -Q r:t
-
/
I
I I... / rI ,6
.5
z·
"tl " tl
t6
~
!
..
/
Sinusoidal 2 Hz
5 X 10- 6
8
4l
10-6
2 X 10- 6
10
15
20
30
40
50 60
aK. MPa • m,/2
Load/time waveforms and FCP rates for specimens of Inconel 718. Top: Various forms of cyclic stress fluctuations used at 550 °C (1025OF)at a frequency of5X 10- 2 Hz. Bottom: FCP rates at 550°C under sinusoidal, triangular and square loads.
To separate out the possible effects of creep or environment from deformation mode, the authors used triangular and square wave shapes, as shown in the top graph. The data obtained using the triangular wave at 2 Hz were the same as the data obtained in other tests using the sine wave at the same frequency which resulted in the lowest FCP rate. The effect of loading at the same rate but imposing a lO-second hold time at maximum load was to increase the FCP rate only slightly, as shown in the lower graph.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 295
11-22. Inconel 718: Fatigue Crack Growth Rate in Air vs Helium ~K, ksi •
iny2
10 20 3040 6080 5 X 10-1 1'"T'"--"""--T"-"""-T""T""1'""'I""l
Air
•
=
o
= He
10- 3
.,
u
~
--EE
10- 2
z'
--.,'.. "tl
Gl
U
> u ...... .5
"tl
.....
10- 4
.I:
i0
..
--. "tl
0>
... ., ..
Z'
"tl
.¥
u u
10-3
::;,
0> .;::;
u..
10-5
10-4 ......._ _L - - J ' - -........I-l'-'-.......... 10 20 40 60 80100 ~K,
MPa . m1/2
Fatigue crack growth rate data for Inconel 718 in air and in helium. Frequency, 0.1 Hz. Temperature, 650°C (1200 OF). Here it is evident that crack growth rate at constant temperature is lower in inert gas.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 287
293
294
11-23. Inconel 718: Effect of Environment on Fatigue Crack Growth Rate ~K, ksi • in. 1/ 2
10
20
30 40 60 80
5 X 10-1 10- 3
.,
u
~ E E
--
10-2
.,
z·
U ~
~ III
"'C
....,'
10- 4
E
--.5 z· -"'C
.t:
i
III
e "'u..."
"'C
CI
III U
10-3
.,
:::l
CI
'';::; III
u..
10- 5
10- 4
• =
He
+ 0.5%
0=
He
+ 5% S02
H 2S
L-_----I_~---L___L.....I.....L.J...L.J
10
20 ~K,
40
60 80100
MPa • m 112
Fatigue crack growth rate data for Inconel 718 in helium + 0.5% hydrogen sulfide and helium + 5% sulfur dioxide. Frequency, 0.1 Hz. Temperature, 650°C (1200 OF).
From the data above it becomes obvious that fatigue crack growth rate increases greatly in aggressive environments compared with exposure to helium alone.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 288
11-24. Inconel 718: Fatigue Crack Growth Rate in Air Plus 5% Sulfur Dioxide AK. ksi • in. ' / 2 10
20
30 40 60 80
5 X 10- 1 Air
+ 5% S02
10- 3
II>
Q > u
--
10- 2
E E
z·
--'" "C
II>
Q
>
"C
i ...'"
10- 4
~
i
~ .S
z
~
e
'"
"C
." ~
u
eu
10- 3
II>
:s
." .;;
'"
II..
10-4
'--_ _.J...._-'---'--'-...L..J.....L..L..J
10
20
40
AK, MPa •
60 80100
m' /2
Fatigue crack growth rate data for Inconel 718 in air + 5% sulfur dioxide.
(The effect of air plus 5% S02 was similar to the effect of air alone.) It was observed that in the helium atmosphere, which was used to establish a baseline, cracking was generally transgranular with well-defined striations. In the air, oxygen-bearing and sulfur-bearing environments, the crack path changed from transgranular to intergranular, indicating that an important effect of the environment was to degrade the boundary strength by mechanisms that were not clearly defined. It was suggested that oxygen diffusion along grain boundaries and localized oxidation may have occurred. Another very important observation was that the effect of a given environment on FCP could not be predicted on the basis of unstressed exposure tests. The attack on the surfaces of unstressed specimens in aggressive S02 environments was minimal, but the S02 environments caused substantial increases in FCP.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 289
295
296
11-25. Inconel 718: Fatigue Crack Growth Rate in Air at Room Temperature
Spec. 1290" / (CHT) 'f
/
~ E
,
:
III
Spec. 158 & 803'" (CHT) /
v'
10-3
E
/
(!.
.. ,I:!> 0
/ , .. I
Z' ~
"~.'"
Inconel718 Tested in air at 24°C (75°F) 500 < f < 600 cpm, R = 0.05
Conventional heat treatment
I:!> Spec. 1290, heat I 0 Spec. 158 } heat II V Spec. 803 0 Heat III 20
40
60
80
0
Modified heat treatment
I:!> Spec. 1283, heat I 0 Spec. 253, heat II
I I
(b)
20
40
60
80
Stress-intensity factor range, ~K, MPa • m 1/2 Fatigue crack growth rate behavior ofInconel718 tested in air at 24 ° C (75OF).CHT= conventional heat treatment. All testing was done at R = 0.05 and at a frequency of 0.67 Hz.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,..in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH. 1982, p 276
11-26. Inconel 718: Fatigue Crack Growth Rate in Air at 316 °C (600 OF)
Spec. 210 (CHT)
-:
y "IJ./ .. ,
Spec. 1294 (CHTl
.I IJ./
/ ,~ Inconel718 Tested in air at 316°C (600°F) f = 40 cpm, R = 0.05
t:J.
..
..I1J.4' ,0
/ IJ.I 0 .. IJ. .,
/ /
./
Conventional heat treatment
IJ.
IJ.
Spec. 1294, heat I
o Spec. 210, heat o Heat III
(a)
Modified heat treatment
10- 6 L..-_ _--I._--'_....L--L.-L-'-...L...J-'40 60 80 20
Spec. 1282, heat I
o Spec. 254, heat
II (b)
II
.L..-_....L--''--J.-........--L....J-I
20
40
60 80
Stress-intensity factor range, Ll. K, MPa . m 1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 316°C (600 OF). CHT = conventional heat treatment. All testing was done at R = 0.05, and at a frequency of 0.67 Hz.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 277
297
298
11-27. Inconel 718: Fatigue Crack Growth Rate in Air at 427 °C (800 OF) 10-1
...----r--,-__r-..___..___.....-r::I:~--__r--..___.....,..____.____.___r.... T.:I Inconel718 Tested in air at 427°C (800°Fl f = 40 cpm, R = 0.05 Conventional heat treatment
10- 2 D.
o
Modified heat treatment
Spec. 1291, heat I Spec. 162, heat II
D.
o
Spec. 1286, heat I spec. 255, heat II
I
:1 Spec. 162 (CHT1,,/1
/'1
/.~
10-3
Spec. 1291",: ~ (CHT) lID.
0 0
0
.. I
II .. I II
10-4
.. I
.II
// 10-5 L...-
L . - _ - ' - - - ' - _ . L - . L -...................
20
40
60
80
0
<9
----'_ _.L---L----'----''---L...............
20
40
60 80
Stress-intensity factor range, AK, MPa • m1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 427°C (800 OF). CHT = conventional heat treatment. All testing was done at R = 0.05 and at a frequency of 0.67 Hz.
Source: Stephen D. Antolovich and J. E. Campbell. "Fracture Properties ofSuperalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH. 1982, p 278
11-28. Inconel 718: Fatigue Crack Growth Rate in Air at 538 °C (1000 OF) 10- 1
1:""""----,--...-~-,.._r_1r_T"T"']r:__--__r--.___r-,.._r_1r_T"""1"":I
Inconel 718 Tested in air at 538°C (1000°F) f = 40 cpm. R = 0.05 10- 2
10- 3
10- 4
10- 5
Conventional heat treatment
Modified heat treatment
Spec. 1288, heat I Spec. 165, heat II heat III
Spec. 1284. heat I Spec. 251} heat II Spec. 250
'--_ _----'C--_.1.--L_.l.-.L.....JL..-L....L.J'--_ _----'_ _-'------'_.l.-.L:--JL.....L~
20
40
60
80
Stress-intensity factor range, LiK. MPa . m 1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 538°C (1000 OF). CHT = conventional heat treatment. All testing was done at R = 0.05 and at a frequency of 0.67 Hz.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of'Superalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 279
299
300
11-29. Inconel 718: Fatigue Crack Growth Rate in Air at 649°C (1200 OF) 1 10- ~---r---r-'---'-"'T1-r:r-----r----,--,--r""'T"r-r""'] Inconel 718 Tested in air at 649°C (1200°F) f = 40 cprn, R = 0.05
V·'
/
Spec. 1289 (CHT) •
10- 3
./ ./
/
!SJ
/ 10- 4
Conventional heat treatment b.
Spec. 1289, heat I
o Spec. 156, heat
o 20
Modified heat treatment II
Spec. 1281, heat I Spec. 252, heat II
heat III 40 Stress-intensity factor range, l1K, MPa . m 1/2
Fatigue crack growth rate behavior of Inconel 718 tested in air at 649°C (1200 OF). CHT = conventional heat treatment. All testing was done at R = 0.05 and at a frequency of 0.67 Hz.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties ofSuperalloys."inApplication of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH. 1982, P 280
11-30. Inconel 718: Fatigue Crack Growth Rates at Cryogenic Temperatures
•
22° C (72° F) } -78° C (-108° F) 2.54 cm 'V -196° C (-320° F) thickness o -269° C (--452° F)
l>
Q)
U
e
10- 3
22° C (72° F) 0.51 cm thickness
o
EE z
~
"0
~f ~
.r:
~
e Cl
.:.!
~
CJ
10- 4
Q)
:I
Cl .;:;
u..
5
10
50
100
Stress-intensity factor range, AI<, MPa • m 1/2
Fatigue crack growth rates of Inconel 718 forged bar at an R ratio of 0.1 and a frequency of 20 Hz. Heat treatment: 980°C (1800 OF) '% h, AC; double aged 720°C (1325 OF) 8 h, FC to 620°C (1150 OF), hold 10 h, AC. At the constant frequency the effect of higher temperature is to increase the FCP rate.
Source: Stephen D. Antolovichand J. E. Campbell, "Fracture PropertiesofSuperalloys,"in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 298
301
302
11-31. Inconel718 and X-750: Fatigue Crack Growth Rates at Cryogenic Temperatures
Inconel 718 (Ref 8.44) ~ 22 to -269°C (72 to -452°F)
Inconel 718 (Ref 8.49)
•
• 22°C (72°F)
Inconel X-750 (Ref 8.48) ~ 27 to -269°C (80 to -452°F)
5
10
50
100
Stress-intensity factor range, ~K, MPa • m 1/2
Fatigue crack growth rates for Inconel718 and Inconel X-7S0in the subzero temperature range.
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, along with room temperature FCP data for Inconel 718 from Shahinian et al. The FCP data for these two alloys overlap in the t!K range shown. 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. However, results of FCP tests depend on both melting practice and thermomechanical processing.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OB, 1982,P 300
11-32. Inconel X-750: Effect of Temperature on Fatigue Crack Growth Rates
• 22°C (72°F) -196°C (-320°F)
CIl
u
!E
A
10- 3
o -269°C (-452°F)
E Z
~ "tl :!l' E
ie
en
~
CJ
b
10-4
CIl
5, .~
'"
II..
100 Stress-intensity factor range, llK, MPa . m 1/2 Fatigue crack growth rates of Inconel X-750 at an R ratio of 0.1 and at frequencies of 20 to 28 Hz. Heat treatment: solution treated and double aged. Within this frequency range, the effect of higher temperature is to increase the FCP rate.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 299
303
304
11-33. Jethete M 152: Interrelationship of Tempering Treatment, Alloy Class, and Testing Temperature With Fatigue Characteristics ksl 100r------------------,
o
...:::>~
'" ~
0.6
)(--------x
0.4 =ksi 100
80
0-_0______
50
i!:
...~
0"
...
G
~
o
x
x tempered at 510 F(300 Cllor one hour to alensile slren&lh 0labout205 ksl
0_
o temperedal1200 Fl650 ClIorone hour toa tensilestreneth abouI150ksl
40
20 LONGITUOINAL SPECIMENS ROTATING·8EAM TESTS FOR \0' CYCLES
0'
200
400
600 I
50
60
0:
Ii;
1000 F
800 1
450
°O!----..!:--....,-k".--""*';---=----.od;;--__=! 1200 I I
200
300
400
600
TEST TEMPERATURE
Left: Interrelationship of prior tempering treatment and testing temperature with limiting fatigue stress, and with fatigue ratio for Jethete M152. Right: Influence of alloy class and testing temperature on fatigue strength for the same alloy.
Source: J. Z. Briggs and T. D. Parker, "The Super 12%CrSteels," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed.. American Society for Metals, Metals Park OH, 1979, P 123
C
11-34. Lapelloy: Interrelationship of Hardness and Strength With Fatigue Characteristics ksi
9o,------------..----,
CLASS II (Lapelloy)
90,---------------,
2000 F C1 095 Cl sail quench, marlemper 650 F(345 C) + lemper 1150/1600 F(620/B70 C) CANTILEVER ROTATING·BEAM TESTS mechanically polished,
BO
rms 2.5/4.0 mlcrelnehes
alhersurface Irealmenls, o Including surface lolline.
rough e:rinding andlough machining. rms5.5/40 microlnches
I
50
70
95% conlidencelimits
50
4~'='00,..----,-!;"....----:-!-::-----:*"--""IBO ksi
Left: Relation between surface hardness and mean fatigue limit for Lapelloy. Right: Relation between tensile strength and mean fatigue limit for the same alloy.
Source: J. Z. Briggsand T. D. Parker, "The Super 12%Cr Steels," in Source Book on Materials for Elevaled-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 123
305
306
11-35. Mar-M200: Effect of Atmosphere on Cycles to Failure
10 8 •
o
•
o 6
AIR DRY AIR WET AIR VACUUM PREOXIDIZED SPECIMEN. VACUUM TESTED
1
10
I
10 8
101
w
o z
'"a:
1
V> V>
w
a:
lV>
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.
Convergence of air and vacuum data was noted for AISI 3I6 steel at 1090 K, and a crossover of the air and vacuum curves occurred for nickel, where it was suggested that oxide in cracks could prolong life in air at low stresses. Crossovers have also been seen in a ferritic stainless steel and a Nil Cr alloy in the range 875-1025 K, where tests in purified argon gave shorter endurances than those in air, impure argon, or sulfur dioxide. Also, in single crystals of the alloy Mar-M200, air endurances were less than those in vacuum at room temperature whilst the reverse was true at high temperature (above). A thin oxide film, formed during testing, suppressed surface crack initiation, but oxide formed during pre-exposure did not.
Source: R. H. Cook and R. P. Skelton, "Environment-Dependence of the Mechanical Properties of Metals at High Temperature," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, P 81
11-36. Mar-M509: Correlation of Initial Crack Propagation and Dendrite Arm Spacing 5.0
..
3,0 u>.
U U
>. u
u
4.5
N""-
c
E E
'fg
'1'0
:il~ 4. 0 :£ ...
:il~
N""-
-
u 2.5 1;
or:
or:
c
c
0
0
iii
0\
:!. 0
iii0\
3.5
:!. 0
~ ij
......
~
u
u
;;;
;::
c
... ...... u
no
2.0
3,0
2.5 L-.. 20
~c
.L-
40 Dendrite Arm Spacing
...r........
.....
60 1~1
Correlation between the initial crack propagation rate and the dendrite arm spacing for Mar-MS09.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys,' in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, p 338
307
308
11-37. Mar-M509: Correlation Between Number of Cycles Required to Initiate a Crack and Dendrite Arm Spacing 300l
r-----~---____r----
......---__r---_,
~ ~
u
...
... 2000
;§ c
-a......
- - - PresentStudy
~
u '0
~ e :::J
IlXXl
z
OL.-_.L.-_--L.
o
20
-'-
40
---'
-L..
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.
Source: Eric Bachelet and Gerard Lesoult, "Quality or Castings or Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984. P 337
11-38. MERL 76, P1M: Axial Low-Cycle Fatigue Life of As-HIP'd Alloy at 540°C (1000 OF) 100 90 00
80 0 'in
~
70
KT
0 =
1.0
m'
fA
60
50 40 30 Life, cycles
Axiallow-cycIe fatigue life ofas-HIP'd P/M alloy MERL 76 at 540°C (1000 OF) at notch severities as indicated.
Source: J. H. Moil, V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications," in Powder MetallurgyApplications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, p 275
309
310
11-39. Nickel-Base Alloys: Effect of Solidification Conditions on Cycles to Onset of Cracking 1000
F"t Solidification (Condition f)
SlolI Solidlficotion (Condillon S)
...
~... ......:> Q
......... 100 ..J
0.
r
... Q
...'" ... ..J
~
U
s ~ ...a: ......r
10
Q.
... Q
...a: CD
r :z :>
M21 71) 1I~71a IN!U M21 713 IN738 '''939 lC LC lC lC "~_ _..L.----1_......L_..L.._..L-_---1L.-_-L..._"""'---1_......L_-L...
(;;:s;J
o
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 2.5 mm in several nickel-base casting alloys.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superal1oys: Source Book, Matthew J, Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, P 339
11-40. Rene 95 (As-HIP): Cyclic Crack Growth Behavior Under Continuous and Hold-Time Conditions Stress-intensity factor range, L1K, ksi • in. 1/ 2 6
8
10
20
40
60
80 100 10-2
10- 1 Stress-level dependence Ql
Ql
U
10-3
>u
-<;
E E
z'
Z'
15-minute hold time at maximum tensile stress
-IV
..
"C
Ql'
-"C IV
.....
"C
Ql'
IV
f!
10- 4
..c
~
Cl
~
>u '<, .~
10-2
"C
...0
U
10- 3
..c
~ 2Cl
~
u
u
f!
f!
u
U
Ql
::3
Ql
Cl
::3
'';:;
Cl
'';:;
IV
10- 5
IV
U.
U.
10- 4
'--.1...--'-_---1._ _........." --_ _- ' -_ _...1-_"'-----'-1
8
10
20
40
60
80
10- 6
100
Stress-intensity factor range, L1K, 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).
The effect of environment need not always lead to more rapid crack growth. It has been proposed that oxidation products could form in the crack tip region and prevent crack resharpening during the unloading portion of the cycle. If the stresses are sufficiently low, the oxidation products in the crack tip region will not be cracked and, in some systems, an elevation of the threshold might occur. Such effects would be pronounced at high temperatures and long hold times and have actually been observed in Rene 95, as shown in the above chart. Once the stress intensity is high enough to crack the oxides, the rate of crack growth would be expected to increase due to the severely degraded region in the crack tip zone.
Source: Stephen D. Antolovich and J. E. Campbell. "Fracture Properties of Superalloys. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood. Eds.. American Society for Metals. Melals Park OH. 1982. P 284
311
312
11-41. Rene 95: Effect of Temperature on Fatigue Crack Growth Rate Testing temperature, • F
100
10_2 20
1000
..
2000
10-4
U
{
.5 z· -.r
z·
....
~
~
.!
E
... ... i'
10- 3
:i l!
~
~ e 11'"
e
11'"
.." l!
10-6
:::J
..~ :::J
'"
.~
...
'"
...'= 10- 4
o Testing tempereture, ·C
Effect of temperature on fatigue crack growth rate at constant for Rene 95.
bJ(
That the effect of environment can be large may be inferred from some low-cyclefatigue studies of Rene 95 in which surface and subsurface cracking was observed at comparable strain ranges and defect sizes. As expected, the life of the subsurface crack was much greater than that of the surface crack, leading to the hypothesis of a strong environmental effect. This possibility is considered in more detail in an analysis of FCP properties of Rene 95. The FCP rate was plotted as a function of temperature for a given !:J.K range, as shown in the above chart. It is noteworthy that there is a minimum in the FCP rate at all !:J.K levels except 22 MPa·M 1/2 (20 ksi-in.b"), where the data are at least suggestive of a minimum. Because any environmental interaction is thermally activated, the crack growth rate at a given !:J.K level and frequency may be written as:
da
dN = Aexp - Q(!:J.K)/ RT
where A is a constant and Q(!:J.K) is the apparent activation energy.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys, "in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 282
11-42. 8-816: Effect of Notches on Cycles to Failure at 900°C (1650 OF) 40 ,----,r----,r--....,...-...,..--r---r---r--..---r---,...---. 280 ltJ
a..
~ vi ~
~
30
210
"-
"-
'Iii
'n. ...........
E :J
E
.~
~
vi ~
20
A::;oo
0-_-0... K,
= 3.4
----__
Reversed stress fatigue
0 r-v
--00"--
106 Number of cycles SoN diagram for 8-816 heat-resisting alloy tested at 900°C (1650 OF), notched (broken curve) versus unnotched (solid curve).
Source: High-Temperature Fatigue, p 245
140
'Iii E :::J E
'x ltJ ~
313
314
11-43. Udimet 700: Fatigue Crack Growth Rates at 850°C (1560 OF) AK, ksi • in. 1/ 2
10
40
20
60 80100
10- 3
10- 2 Q)
u
Q)
u
> t.l
--EE
> t.l
--.E
z·
z'
"tl
~
'<,
'"
10- 4
"tl
...~ Q)'
.t:
~
01
-" t.l
...e
Q)'
s:
e
'"
"tl
~
e 01
10-3
-"t.l
~
~
t.l
t.l
.'"
Q)
Q)
...
01
.01
u,
'"
u,
::J
::J
10- 5
Stage I 0
m = 16
KN
10- 4
Ib
c 582 1310 v 711 1600 01067 2400 l> 1244 2800 160 3600
°
20
10- 6 40
60 80 100
AK, MPa • m 1/2 Crack growth rates in terms of stress-intensity factor range for Udimet 700 at 850 0 C (1560 0 F). Crack growth rates for this alloy are greatly accelerated by increases in temperature.
Source: Stephen D. Antolovich and J. E. Campbell. "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for Metals, Metals Park OH, 1982, P 285
11-44. U-700 and Mar-M200: Comparison of Fatigue Properties I
- - COLUMNAR GRAINE\i AND SINGlE CRYSTAl
-- .. -- --.
M~R.:-,~OO
w
Z
10-2
-c
1700 0 F
-c
-
a: Z a:
1'-.
l-
V)
..... -(
I-
-. -- ..
I-
w- 2
oI-
I
I
1-..
C>
I
- - - CONVENTIONAllY CAST MAR.M200 I I I --·-WROUGHT POIYCRYSTAlliNE UDiMEl 700
.~
I':::--
--_.~~
-
I
I
r--
-.
'--.
I-.. r--..,.--. .1---•. - r-
~
I
I
!--.
'.
1400 0 F
, 103
CYCLES TO FAILURE Comparison offatigue properties at 760 0 C (1400 0 F) and 925 0 C (1700 OF) for a typical wrought nickel-base alIoy (U-700) with conventionalIy cast, directionally solidified and mono crystal , Mar-M200.
Source: Francis L. Versnyder and M. E. Shank, "The Development of Columnar Grain and Single Crystal High Temperature Materials Through Directional Solidification," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 358
315
316
11-45. Waspaloy: Stress-Response Curves
0
a. 1200
::iE Ul
c
STRAIN CONTROLLED
1100
:::l
t-
:J a. ::iE
«
V) V)
Ul 0::
t-
WASPALOY
V)
10
100 N, CYCLES
1000
Stress-response curves for Waspaloy having nonshearable precipitates.
During aging of precipitation-hardenable alloys, the coherent precipitates grow, and accommodation strains build up. At some point the energy associated with the accommodation strains exceeds that necessary to create a precipitate-matrix interface, and the precipitates become partly incoherent. This is accompanied by a change in precipitate-dislocation interaction from one of shearing to that of dislocation looping or bypassing the precipitates. Since the reasons for strain localization have been removed, deformation becomes more homogeneous. Local softening is thus prevented, and the cyclic-response curve shows hardening to saturation, or to failure, as illustrated above.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 217
11-46. X-40: Effect of Grain Size and Temperature on Fatigue Characteristics
~ %
} -'---} ----
2
Z .... <
at 750 DC at 650 DC
~ 0.6 1Il L!J
z
~
0.5 0.4
z< 0.3 0::.
Small
w
I-
...J
-e
0.2
-.... 0.1 2 100
5
2 1000
5 10.000
CYCLES TO FAILURE S -N curves for X-40 showing effects of grain size and temperature on fatigue characteristics of this alloy.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH. 1984, p 335
317
318
11-47. Cast Heat-Resisting Alloys: Ranking for Resistance to Thermal Fatigue
IlJ
a::
:J
='300
it
o F-200
~150
...J
~IOO
ILl
~ 50 u, 40
o
ffi
30
z~
20
ILl
15
ILl
10
m
~ a::
~
10 15 20 30 40 50 100 200 300 CYCLES TO FAILURE OF INDIVIDUAL TESTS
The design of components that are subject to considerable temperature cycling must also include consideration of thermal fatigue. This is particularly true ifthe temperature changes are frequent or rapid, and nonuniform within or between casting sections. Fatigue is a condition in which failure results from alternating load applications in shorter times, or at lower stresses, than expected from constant-load properties. "Thermal fatigue" denotes the condition when the stresses are primarily due to hindered expansion or contraction. Good design helps minimize the external restraint to expansion and contraction. Rapid heating and cooling may, however, impose temperature gradients within the part causing the cooler elements of the component to restrain the hotter elements. Finite-element computer analysis has shown that, for some industrial applications, these thermally induced stresses may exceed those resulting from the mechanicalloads. An example of results from thermal fatigue data is presented above. This graph offers a ranking of many cast heat-resistant high-alloy grades relative to their resistance to thermal fatigue. Such rankings are indicative of general alloy properties only because most thermal fatigue tests are based on an arbitrary set of experimental conditions rather than on their fundamental material behavior. Nevertheless, such test results have been useful in considering alloy selection questions, and in identifying the superior thermal fatigue resistance of nickel predominating grades and the good performance of some HH type compositions.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 19-7
12-1. Corrosion-Fatigue Properties of Aluminum Alloys Compared With Those of Other Alloys
500,.------typical
corrosion
-.,
fatigue strength
(sea water , N • 108 cycles ambient temperature
t
I
70
R· -1 )
60
400
';:;'"""'
-€. z
duplex staness steels, titanium alloys 50 (e.g. n -6A1-4V)
~
0
300 40
J:.
0. c:
e
-F
200
~
.~
~
~
0
'iii
.>< L-J
0
t
iii
s
t
,.....,
100
30
.!j
nickel alloys (e.g. aJloys 600 and fn»
r.. .·-/ .
.S!'
~ 20
ferritic _ stainless martensilic_ \ steels
copper - nickel alloys ,carbon steels, I bw alloy aluminum m
iii
.~
e
8 1)
0
As shown above the corrosion-fatigue strength of bare aluminum alloys is superior only to that of magnesium alloys. Careful surface protection may bring the corrosion-fatigue strength up into the range of bare stainless steels or copper-nickel alloys.
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, R. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OR, 1980, P 617
319
320
12-2. Comparisons of Aluminum Alloys With Magnesium and Steel: Tensile Strength vs Endurance limit 6.
250
()~ /
':>.o~'b
200
/
"3 /
6.
6./
r0-
/
ll..
~ E::::: 150 :.= (J)
.'=
0) 0)-
u
u
c i:)
moo 100 ~O
"O~
Aged aluminium alloys Non-heat treatable aluminium alloys Magnesium alloys Steels
x
C )( UJ~
50
O~---JL...-_----I.
100
x 0
•
6.
_ _---L_ _-'-_ _....I..-_ _.L......_---l
200
300
400
500
600
700
Tensile strength (MPa) Fatigue ratios (endurance limit/tensile strength) for aluminum alloys compared with those of magnesium and steel.
It is well known that, in contrast to steels, the increases that have been achieved in the tensile strength of most nonferrous alloys have not been accompanied by proportionate improvements in fatigue properties. This feature is illustrated in the graph above, which shows relationships between fatigue endurance limit (S X 108 cycles) and tensile strength for different alloys. It should also be noted that the fatigue ratios are lowest for age-hardened aluminum alloys and, as a general rule, the more an alloy is dependent upon precipitation hardening for its total strength, the lower this ratio becomes.
Source: I. J. Polmear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH, 1981, P 39
12-3. Aluminum Alloys (General): Yield Strength vs Fatigue Strength
•
500
ultimate tensile strength yield strength fatigue strength in air , N - 5 X 101 1 latiguestrength in sea water I N _10 fatigue strength • 8 in river water I N - 10
o o • •
•
R - -1 , .. 60 Hertz I stnooth specimens I ambient temperature
4()()
70
60
+ 50
';:;"'"'
--z E
6
+
.--. 300
•
0
'iii
""
•
'--'
40
0
corrosion fatigue strength 01 aluminum alloys
s:
s:
e;,
e;, c
30
l!! 200 OJ
•
c
l!! OJ
000
o
0CF,air ,
N-5x1Q1
-Orr-----l:I
20
100
000 I
I
I
~§ ~
~
I
~~
II
~ ~~ I
~I
f2,
~
<0 ~ I
Cl)
iij
...
!2 <0, <0 <0 ~ I I , ~ Ie ~ 'Ie R ;:: R ~
~
An analogous conclusion can be drawn from a review of corrosion fatigue tests with smooth aluminum alloy specimens as shown in 'the above graph. Here aluminum alloys are listed in order of increasing yield strength. As the yield strength goes up, so does the ultimate tensile strength, but the fatigue strength in air soon reaches a limit which is roughly the same for alloys of greatly different yield strength. In other words, medium- and high-strength aluminum alloys all have about the same fatigue strength. 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. Thus, corrosion fatigue is still a limiting factor for the application of aluminum alloys.
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 616
321
322
12-4. Comparison of Aluminum Alloy Grades for Crack Propagation Rate
1O- 3 r--- - - - - - - - - - - - - - - - - - - ---, 7050 - T 736 7175 - T 736 2219 - T 851 7079 - T 651 2618- T 6
~
Q)
~
10
5
~ E
'----'
10-6 Z
~
10- 7
10
8
~
typical scatter in experiments
crack orientation L - T specimens CNP, DCB I CT ambient temperature environment air R =0,0-0,1 0,1 -30 Hz
10- 9
I
I
I
---+--------------------4
1).-....
°
10
20
cyclic stress intensity range
50
40 I
.t:.K I
[
60
MN· m- 3f2]
Many commercial aluminum alloys show similar fatigue crack propagation rates in air, as indicated in the above comparison.
Source: Markus O. Speidel, "Aluminum asa Corrosion Resistant Material, "in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloniand D. E. Embury, Eds.. American Society for Metals. Metals Park OH, 1980, p 613
12-5. Alloy 1100: Relationship of Fatigue Cycles and Hardness for HO and H14 Tempers 50r----.,---.------.,---.-------r--r-----r--r-----r--r--,---r--,---r----,
• ~ 30
-----------
LU
Z
o
cz:: <:
:I:
~
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. For instance, in the case of well-annealed Al 1100 (RO), 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, accompanied by pronounced slip-step formation on the surface. On the other hand, in the case of the cold-worked material Al 1100 (R 14), the material will soften in the early stages offatigue (above curves) as the dislocation density, introduced by the cold work, decreases. Slip-step formation in this situation is much less pronounced than it is during hardening, because the initial dislocation-loop length is much shorter. In either case, during this initial rearrangement, the dislocations form a cell structure with individual dislocations of long loop length shuttling to and fro between the cell walls. This latter part of the fatigue life is called the saturation stage offatigue, during which the dislocation shuttling leads to local instabilities, or "extrusions-intrusions," and finally to microcracks, which can be observed after about 25% of the fatigue life has been expended. The micro crack density is about the same for both materials.
Source: O. Buck and G. A. A1ers,"New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 128
323
324
12-6. Alloy 1100: Interrelationship of Fatigue Cycles, Acoustic Harmonic Generation and Hardness 8xlf3
16xlf
3
80
......
~
Cla -e
A
zlA I
4
60 Vl
0
......
Vl ..... z
-c'"
40 oe: « ::J:
~
0
ll.
K.H.
4
0 0
ZO
z~
AI 1100
0 0
80
120
0 160xllY
FATIGUE CYCLES Normalized second harmonic displacement and Knoop hardness as a function of fatigue.
The effects of dislocation rearrangements on harmonic generation within the bulk of the material during fatigue are shown in the above chart. Using 3D-MHz longitudinal waves, the normalized second harmonic amplitude of an initially compressiondeformed Al 1100 single crystal was monitored and found to increase as a function of compression-compression fatigue. At the same time, the surface hardness (Knoop) decreased. Apparently, the dislocation-loop length prior to fatigue was quite short, since the initial amplitude of harmonic generation was small. During fatigue softening, the cell structure that developed (with its individual dislocations within the cells) became quite large, so that a change of the dislocation-induced harmonic generation, A 2d , increased. Application of this technique to highstrength aluminum alloys failed, however, apparently because of an immediate rep inning of the long loops by interstitials in this alloy.
Source: O. Buck and G. A. Alers, "New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 131
12-7. Alloy 2014-T6: Notched vs Unnotched Specimens; Effect on Cycles to Failure
50
~ 40 )(
N.
c:
~ 30 VI
VI .... cr: :n
20
10
Effect of notch on fatigue of 2014-T6. As is true for most alloys, notches greatly reduce the fatigue properties of aluminum alloys.
Source: P. C. Varley, The Technology of Aluminium and Its Alloys, Butterworth & Co. Ltd., London, England, 1970, p 43
325
326
12-8. Alloy 2024-T3: Effect of Air vs Vacuum Environments on Cycles to Failure
..
40
~ 35
Ii Q)
J; 30 Cl
c
.- Ultrahigh Vacuum 0- Air
, ~
'"
.........
..........
<,
<,
"0
; 25
~,
<,
.........
c
En'durance Limit
15 s 10
I I I I III"' 2
4
6
8 10'
2
4
6 8 10
. "~
...
-
J 7
2.5 x 10
l/)
~~
I
-3
-
)(
~
-3
•
'~ r""- ...
III
E ::J ,§ 20
r-
- 3.5 x 10
2
1.5 x 10- 3
3
Number of Cycles-to-Failure The effects of air versus vacuum on the fatigue life of a 2024-T3 aluminum alloy.
For most materials, environment appears to be most effective early in the crack-growth process, with little or no effect at high crack-growth rates. Additionally, the majority of S- N curves diverge at decreasing stresses, the increase in fatigue life caused by vacuum becoming greater at lower stresses. In contrast to this behavior, however, aluminum and aluminum alloys have been shown to exhibit conflicting results. For example, 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- N curves at low stresses, the effect of environment apparently becoming less important at decreasing stresses, as shown in the above chart.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 337
12-9. Alloy 2024-T4 Alclad Sheet: Effect of Bending on Cycles to Failure 400 350
rf.
:E VI' VI
...e
(/)
50
300 40 'w ...:
250
~.
30
200 150
-----2 1
Stress ratio. 0.1
100
0.01
....::.4"-3 0.1
e
en
20 10
Millions of cycles to failure
Effects of bending on fatigue characteristics of aluminum alloy sheet.
For the data here, sheet 1.02 mm (0.040 in.) thick was annealed, solution heat treated and quenched, and then fatigue tested. The sheet represented by curve I was not bent. All other sheet was bent 90° in the annealed condition. Flattening (unbending) was done in either the annealed condition (curve 2) or the solution heat treated and quenched condition (curves 3, 4 and 5). Details of bending and flattening were as follows: (I) Not bent. (2) Bend radius, 3.18 mm (Ys in.); flattened in annealed condition. (3) Bend radius, 3.18 mm (Ys in.); flattened in quenched condition after 3 days of storage at-18 to-12°C (0 to IOOF). (4) Bend radius, 3.18 mm (Ys in.); flattened in quenched condition after 14 days of storage at -18 to -12°C (0 to 10 OF). (5) Bend radius, 1.59 mm (1/ 16 in.); flattened in quenched condition after 3 days of storage at - 18 to - I2 ° C (0 to IO° F).
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 35
327
328
12-10. Alloy 2024-T4: High-Cyclevs Low-Cycle Fatigue 40 ~
LOW CYCLE t R -0 2 C. RAOIAJlON I -_. HIGH CYCLE (R. - 1 Mo RADIATION I I PANGBORN e I ill I
III
'5 C
]
30
/
/
C::P
/ /
%
5
~
/
20
/ / /
~
Q
w
~a: a: 8
A /,
/,
/
/
/
10
/ /
0 0
0.5
1.0
FRACTION OF LIFE In/n,1 Dependence of 13 on n/nior low-cycle fatigue and bulk properties of high-cycle fatigue of AI 2024.
After correcting for the difference in initial /30 values, it can be seen in the above diagram that the two fatigue processes, although radically different in strain history, exhibit similar behavior throughout most of the fatigue life.
Source: Sigmund Weissmann and William E. Mayo, "Determination of Strain Distributions and Failure Prediction by Novel X-ray Methods," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf, Eds., American Society for Metals, Metals Park OB, 1984, p 195
12-11. Alloy 2024-T4: Relationship of Stress and Fatigue Cycles
32 30 28 u
26
~
0
.....0
24
VI
-;'"
c::
22
E
I~
20
sc;
18
..,.z
.c::
e'"
16
0'
«'" >
14
,,
12 : 5,000
10,000
15,000
20,000
Number of fatigue cycles, N Dependence of 13 on number of cyclesN at various stress levels of AI 2024-T4.
Here is shown that for the maximum stress of241 MPa with R
= 0.1, the 13 value increased during the first several hundred cy-
cles. This was more pronounced for the surface grains (Cr radiation).
KO'I
Source: Sigmund Weissmann and William E. Mayo, "Determination of Strain Distributions and Failure Prediction by Novel X-ray Methods," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf, Eds., American Society for Metals, Metals Park OB, 1984, P 194
329
330
12-12. Alloy 2024-T4: Dependence of the Average Rocking Curve Halfwidth 13 on Distance From the Surface 20..--------------------, 19 18
~-_:_------,O----..."....----_1
o
ICQ.
o
b
14 13
_ _ _ _ _ ~o (inillal ~alf.i~_~_)
_
12
IIL..-_ _..........
o
50
.L--_ _- - ' -
100
150
....l...- _ _---J
200
250
I'm
Dependence of the average rocking curve halfwidth 11 on depth distance from surface for different fractions of corrosion fatigue lives, N F' of Al 2024-T4.
X-ray rocking curve measurements were carried out as a function of depth distance from the surface, and typical results ofthe dependence of 13 on depth distance for an alloy cycled with a = 276 MPa, corresponding to the static yield stress, are shown above. It may be seen that the minimum 13 values at the surface layers were larger than those in the interior. The 13 values declined up to a depth of about 50 Mm from the surface and subsequently retained a plateau value throughout the interior of the specimen for each fraction of the life.
Source: Sigmund Weissmann and William E. Mayo, "Determination of Strain Distributions and Failure Prediction by Novel X-ray Methods." in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf. Eds., American Society for Metals, Metals Park OH, 1984, P 193
12-13. Alloys 2024 and X2024: Effect of Alloy Purity on Cycles to Failure +
260 tV a...
0
:2:
CIl CIl
~
220
CIl
Ol C .~
C
~ 180
+
140 ~ 105
---'106
o ---'----'
Cycles to failure Effect of reducing the concentration of submicron particles in an AI-Cu-Mg alloy. X2024 is a highpurity version of the commercial alloy 2024.
The disappointing fatigue properties of age-hardened aluminum alloys are also attributed to an additional factor, which is the metastable nature of the metallurgical structure under conditions of cyclic stressing. Localization of strain is particularly harmful because the precipitate may be removed from certain slip bands, which causes softening there and leads to a further concentration of stress, so that the whole process of cracking is accelerated. The fatigue behavior of age-hardened aluminum alloys would therefore be improved if fatigue deformation could be dispersed more uniformly. Factors which prevent the formation of coarse slip bands should assist in this regard. Thus it is to be expected that commercial-purity alloys should perform better than equivalent high-purity compositions because the presence of inclusions and intermetallic compounds would tend to disperse slip. This effect has been confirmed for an Al-Cu-Mg alloy, and fatigue curves for commercial-purity and high-purity compositions are shown in the above S-N diagram. Here the superior fatigue behavior of the former alloy arises because slip is more uniformly dispersed by submicron dispersoids such as MnAI 6 •
Source: I. J. Po 1mear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH, 1981, P 40
331
332
12-14. Alloys 2024 and 2124: Relationship of Particle Size and Fatigue Characteristics Ii K, ksi • in. 1/ 2
3
5
8
10
15
20
10- 3
10- 2 Q)
Q)
u
> u
--
E E
10-4 10- 3 10-5
'C
10-6 10-5
e
.. .'" ~ 0
en
10- 7
en
'';:::;
Q)
'"
'0.
u
c,
:l
en
'~
t
'C
.~ Q)
e, UI
o
.. -~
'u
U.
10-6
1......
.:.: o Q)
Q)
:l
'"
....
...
en
U.
'C
'" .s:
~ 0
u
--'"
Q)"
10- 4
s:
.:.:
--.5 'C
'C
Q)"
> e
z·
Z
--'" ..... ...'" . '".
u
Q)
1
"---'~_ _"""""_ _"""""---''--_''''''''''_'''''''''---' 10- 8
20 Stress-intensity factor range, LiK, MPa • m 1/2 Comparison of typical particle sizes in aluminum alloys with crack advance per cycle on fatigue loading.
The above graph represents Staley's work in summarizing the role of particle size on fatigue crack growth in aluminum alloys,
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys." in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OK 1982. P 191
12-15. Alloys 2024-T4 and 2124-T4: Comparison of Resistance to Fatigue Crack Initiation
Mechanically Polished R=-I
E
z
~
2.6
0
0
~
~
Z
A
~
0
2.4
...
in
co
..r:::
CIl
0
200
fJl
"0 0> 0
300
fJl fJl Ql
AA
0
z
400
N
...
Ql fJl
Nij =Cycles to ~ 15 fLm Crock
A
fJl fJl Ql
in
600 500
~ <,
700
2024 T4 oNjj 2124 T4 ANjj
2.8
0
co
2.2
A Ao
...J
0
2.0
102
103
105
..r:::
AA
0
"0
z 106
107
Cycles to fatigue crack initiation for specimens of aluminum alloys 2024-T4 and 2124-T4 versus stress at notch base (computed using Neuber stress-concentration factor).
The 2124 alloy studied had 1/ 10 the inclusions of the 2024 alloy studied (0.2 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). With 2124-T4, slip-band cracks not associated with inclusions formed at the lowest stress studied. They also formed more easily in 2124-T4 than in 2024-T4 at high stresses, in keeping with the larger grain size. Thus, as shown in the above chart, at high stresses 2124-T4 is less resistant to fatigue crack initiation than is 2024-T4, but it is more resistant at low stresses.
Source: M. E. Fine and R. O. Ritchie, "Fatigue-Crack Initiation and Near-Threshold Crack Growth," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 251
333
334
12-16. Alloys 2024-T3 and 7075-T6: Summary of Fatigue Crack Growth Rates t.K, ksi . in.1/2
10 100 10- 1 r - - - - - - - , - - - - - -....- - - - - - - . , 7075-T6 9 investigations
10- 4 2024-T3 8 investigations
i
>u '<,
10- 5 10-
Note: Bounds defined by mean curves of separate investigations
4
Cl ~
u
E u ~
G>
U
10-
5
.~ z
"C ...... 01 "C
10- 6
Cl
.~
U.
10-7 10- 6 10
100
Stress-Intensity factor range, AK, MPa • m 1/2 Summary of fatigue crack growth rate data for aluminum alloys 7075-T6 and 2024-T3.
Considerable use has been made of the fracture mechanics approach in measurement of fatigue crack growth rates in aluminum alloys. These data have been generated by methods comparable to those of ASTM Method E647 for measuring fatigue crack growth rates. In general, fatigue crack growth rates are found to fall within a relatively narrow scatter band, with only small systematic effects of composition, fabricating practice or strength level, as illustrated by the data in the above chart.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties. of Aluminum Alloys," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerherich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982,P 189
12-17. Alloys 2024-T4 and 7075-T6: Effect of Product Form and Notches
I
'" rn UI
414
_ - -----+---j--+_
345
~---l·~-~~
e,
- --
::E
276
w
207 I
a:
l-
rJ)
138
2024-T4 ---+----1
."
~.,-::'~---+---+---j
I,..... 1'I,--'~-r-''"' ". - ~.....;-"'~_ _+-_--j
I
~ ..
"
'"'-I' ,·t'i:::::~" J -, <,
~
:.:~,.;-.~ c ~-•• -
<,
'q".
~.
"..
69 0
I
I---L.--+~-t---+-
414
I", -,,
7075-16
~----I
345 I----i---"~~.-:-l.c---+----jf----+----j
'"
c,
::E
rn rJ)
w
a:
~'
\
.\ . f~,~~.l,. ~
276
207 1---+---''':-....'
lUI
\
•
''':".":--
--+
,,--
- - -
.~":''-1''''.1 " • t': ':', .... ~'~r, ::""II'jl ~. ,', " .,'\ ~,.
"
"e'::;"',"-j--""'-f'--..
138 ~c-'-cc-:--''''.j,-,''-...,.;t-,§:--,:,r-'- ,,"00.
69
o 1~
:• ,
-
~.,"
:• ~gmg ~~~TE FORGINGS
1'"
....::.:,
• EXTRUSIONS 010 NOT FAil
•
----L
1~
1~
~
1~
---~..!...:'.I,
'---->":" -.#_.s
. _ _ . __ _ _ 1~
10'
! I.
-. --}. "-,.J.'
108
109
CYCLES
Fatigue performance of smooth and notched (K, > 17) rotating-beam specimens from various product forms of 2024-T4 and 7075-T6 aluminum alloys.
Numerous methods have been developed to evaluate response of materials to cyclic deformation. The earliest method was by use of S-Nplots. Typical examples are depicted above. Basic specimens include rotating beam, axially stressed and sheet flexure. Notches have been employed to provide stress concentration, and special specimens have been used to simulate a variety of other conditions. The S- N response is strongly influenced by a number of conditions, including surface condition, stress ratio, and environment. The various alloys differ widely in their response to fatigue testing-specifically, in the number of cycles where a "level out" condition is attained. As shown in the above S- N diagrams, the SoN response for aluminum alloys tends to level out as the number of applied cycles approaches 500 million. Based on SoN data of smooth and sharply notched specimens and of similar tests of specimens designed to simulate joints in structures, the following conclusions have been drawn. From fatigue results for aluminum alloys obtained with smooth specimens . . . rather wide variations can exist without causing appreciable differences in fatigue strengths. . . . When severely notched specimens are used, the effects of composition and temper are even less pronounced and generally are of no practical significance. . . . As in the case of simple notch fatigue tests, there is a lack of significant differences in the fatigue strength of the joints of the various alloys. Despite these laboratory data, users discovered that certain aluminum alloys performed decidedly better than others in service when fluctuating loads were encountered. For example, airframe manufacturers determined that fatigue performance of alloy 7075-T6 was unquestionably inferior to that of alloy 2024-T3.
Source: T. H. Sanders, Jr.• and J. T. Staley. "Review of Fatigue and Fracture Research on High-Strength Aluminum Alloys." in Fatigue and Microstructure. American Society for Metals, Metals Park OH, 1979. P 470
335
336
12-18. Alloys 2024-T351 and 7075-T73XXX: Comparison of P1M Extrusions and Rod 500 r------,r------,-----,.---~---_,_---__r---__, 70 1.35 mm (0.053 in.) 6.43 mm (0.253 in.)
f ---l_+_~=1B --F
+-__-+-_ _
400 1-_ _ <0
0..
::E
Stress ratio: R = 0 Ambient air
vi III
~
-+-_ _--+
t -_ _-1 60
Notch tip radius: e = 0.33 mm 10.013 in.) K, = 3
50
300 f----I------1I-------1I-----j----j-----t-----l ~
III
E :J E
'x
M
<0
::E
E :J E
'x <0 ::E
200 1 - - - - / - - - + - I f - 1 . * - - - I - - - - - t - - - - - t - - - - - + - - - - - - l
20
-:::~~~~~"'F;;;;;:;:;;:::::j::~-__jl--I 100 f -_ _ 2024-T351 rod and bar band -
~"t----I-~ 7075·T73XXX
10
products band
oL -_ _-.l. 10'
103
....L
...L.
10'
105
L -_ _---'
10'
10'
....L_ _~
10'
0
10'
Cycles
Comparison of room-temperature axial stress notch fatigue strength ofP 1Malloy extrusions and ingot metallurgy alloy rod, bar and products, 0 ,X7090·T7E71 in the longitudinal direction; • , X7091-T7E69 in the longitudinal direction; 6., X7091-T7E69 in the long transverse direction; - denotes test specimen did not fail,
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,p 468
12-19. Alloy 2048-T851: Longitudinal vs Transverse for Axial Fatigue 500
'"
70 60 ""'"
Q..
:i:
400
::I"
::I"
I;.
~
'" 300
•
E :::J
E
'x 200 '" :i: 100 103
• R = 0.1
o 0 I;. Longitudinal I • • ... Longtransverse
10 4
I;.
105 106 Numberof stress cycles
50 ~ '" 40 E :::J E
30
'x
'" :i:
20 107
SoN axial fatigue curves for unnotched specimens of aluminum alloy 2048·T851 plate, showing effects of R value and direction upon fatigue properties.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals. American Society for Metals, Metals Park OH, 1979, P 80
337
338
12-20. Alloy 2048-T851: Notched vs Unnotched Specimens at Room and Elevated Temperatures 500
s:
:::E
~
400 ~
Ii: E
300
.~
200
~
10
Unnotched R
r-,
0.1
60 ] 50
24°C 176of)
176°C 13600F)~_
T
~
g E
/120oC 1260 of)
40 ~
-=-
30
:::E
!
20
101 Number of stress cycles
...:::E..
400
g ~ 300
E ~ 200 .~
:::E 100
Notched (K t '" 3.01 R
~ s,
~
0.1
24°C (76 of)
~J
120 °c 1260 of)
176°C 1360 OF)
-
50 ]
- 40 g
30 _ 20
--;;;;1
§ E
"i :::E
Number of stress cycles
S-N curves for unnotched (upper graph) versus notched (lower graph) specimens of aluminum alloy 2048-T851 plate.
Source: Metals Handbook. 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals. Metals Park OH, 1979, P 82
339
12-21. Alloy 2048-T851 : Fatigue Crack Propagation Rates in LT and TL Orientations 1
1
AK. kSi'in. 2
AK. kSI'in.2
10
10
LT crack oriant!tion
TL crack orianLtion
l~3.4
/
" , 10-8
1
10- 5
{.:
G
!
.5 ~.
..:!
rot" -1i ~
10-5
.5
E
~
~
..:!
..:!
0
10-6
M-a.:
M~7.1
61
1
10
G
"iI
~
10- 6
f
1
61
1
10
Fatigue crack propagation in aluminum alloy 2048-T851 plate, showing propagation data for both LT and TL (longitudinal and transverse) crack orientations.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, p 81
340
12-22. Alloy 2048-T851: Modified Goodman Diagram for Axial Fatigue Minimum stress,ksi
s:
60
::;:
~ 1;;
~
E
"E 'x
..
~
E E
"
200
.~
::;:
::;:
Minimum stress, MPa
Modified Goodman diagram for axial fatigue of un notched specimens of aluminum alloy 2048-T851 plate.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 81
12-23. Alloy 2219-T851: Dependence of Relaxation Behavior on the Cyclic Hardening Parameter O~-----r-------,------r-------,r-----...,
o
o
o
0% RH
• 50% RH
-300
L--
o
....J..
10
...!-
20
.l..-
30
---I.
40
...J
50
CYCLES (x 10-3)
The dependence ofrelaxation behavior on the cyclic hardening parameter, (J. (J was varied by changing the relative humidity (RH), which affects the near surface ductility in this alloy. Values used were: (J= 6 X 10.5 for 50% RH and (J= 2 X 10-5 for 0% RH. The cyclic stress amplitude was 0.88 a yleld for both samples.
Source: M. R. James and W. L. Morris. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue." in Residual Stress for Designers and Metallurgists. Larry J. Vander Walle. Ed.. American Society for Metals. Metals Park OH. 1981. P 184
341
342
12-24. Alloy 2219-T851: Effect of Strain Amplitude on the Relaxation of Residual Surface Stress With Fatigue Or-------.-------,r------.------,----..., 0.7" YIELD
o 0.64 a YIELD
-300
.....
o~-----'--------l----...J-.-------l----
10
20
30
40
50
CYCLES (x 10- 3)
The effect of strain amplitude on the relaxation of surface residual stress with fatigue. The symbols are the residual stress value measured by the x-ray diffraction peak shift technique. The solid curves are the predicted mean residual stress values during fatigue.
Surface milling produced the shallowest stress gradient and resulted in the slowest rate of relaxation of the surface stresses. A comparison of measured to predicted values of residual stress during fatigue is made for four "as machined" specimens in the above chart. The residual stress values were measured parallel to the external stress axis. A value of f3 = 0.0004 was used to fit the data for all specimens. Residual stress measurements were also made in a direction transverse to the applied stress axis. Within experimental error, the cyclic relaxation rate was the same as in the longitudinal direction.
Source: M. R. James and W. L. Morris, "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue," in Residual Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 182
12-25. Alloy 2219-T851 : Relationship of Fatigue Cycles to Different Depth Distributions of .Surface Stress O....------r-----r-----,-----r-----,
o
o
•
-300
l--
o
--L
-L-
10
20
•
..L....-
30
----I
40
.....
50
CYCLES (x 10-3)
The relaxation behavior oftwo samples having different depth distributions of residual stress. Note the difference in the peak cyclic stress, a. 0= rolling (10%reduction); f3 = 0.012; a= 0.91 ayield'. = sand blasting; f3 = 0.003; a = 0.71 ayield'
Relaxation of a compressive surface stress requires an expansion of the material normal to the surface. Of necessity, this involves slip at an acute angle to the surface. If the slip does not penetrate the surface, the residual stress cannot relax. Supporting this picture are our observations that the relaxation rate in Al 2219-T851 is more rapid in dry air. It is known that humidity increases the rate of cyclic hardening of a thin (less than I Jlm) layer at the surface. The effect of humidity on relaxation is therefore simply to make it more difficult for dislocations to penetrate to the surface.
Source: M. R. James and W. L. Morris. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue," in Residual Stress for Designers and Metallurgists. Larry J. Vander Walle. Ed.• American Society for Metals, Metals Park OH. 1981, P 183
343
344
12-26. Alloy 2219-1851: Probability of Fatigue Failure
LOG (NUMBER OF FATIGUE CYCLES) Schematic curves of constant probability for failure (actual failure = 100%).
The solid line in the graph represents failure; the dashed lines indicate the percentage offatigue life expended. The exact location of these lines is highly sensitive to the material and its microstructure as well as the influences of environment.
Source: O. Buck and G. A. Alers, "New Techniques [or Detection and Monitoring of Fatigue Damage.t' in Fatigue and Microstructure. American Society for Metals. Metals Park OH, 1979. p 104
12-27. Alloys 3003-0, 5154-H34 and 6061-T6: Effect of Alloy on Fatigue Characteristics of Weldments 300 40 260
30
If. 200
:t!
::;;
g E E
g
160 20
"
..
'x ::;;
E E
.."
'x ::;;
100 10 50
Number of cvcles
The fatigue life of welded joints at high loads varies with the alloy. As the load is decreased, differences disappear until, at about one to ten million cycles of axial loading (R = 0), the fatigue strength of an arc-welded joint is approximately the same regardless of alloy and is 50 to 70% that of the unwelded alloy. Typical data are given in the above graph for three aluminum alloys. Specimens were from 9.5-mm (Ys-in.) plate; weld reinforcement removed; axial loading; R = O.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 195
345
346
12-28. Alloy 5083-0 Plate: Effect of Orientation on Fatigue Crack Growth Rates .1K, ksi • in. 112
5
10- 4
.. u
10- 3
.
u
>
~ -.
-.t.l
E E
10-5
Z
'C -. Cll 'C
.E
Z
'C -. Cll 'C
10- 4
10- 6 T-S 10-5
Compact specimen thickness = 46 mm (1.8 in.) R = 1/3, f = 13 Hz Room temperature, dry air
.1K, MPa . m 1/2
Effect of orientation on fatigue crack growth rates in 180- and 196-mm (7.0- and 7.7-in.) 5083·0 plate.
From the data shown above there is obviously no great effect of specimen orientation on fatigue crack growth rates.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 193
12-29. Alloy 5083-0 Plate: Effect of Temperature and Humidity on Fatigue Crack Growth Rates ~K, ksi • in. 1/2
10- 2
,•
I
,/
RT, moist
air~i
,, ,,,I ,
10- 3 Ql
U
"C
Ql
U
,
~
EE z· ~
10- 4
10- 5
~
.E
z·
~ III
"C
10-4
10-6 Compact specimen thickness = 46 mm (1.8 in.) R = 1/3. f=13&18Hz. T-L orientation
10-5
'--
...L-_--L.
5 ~K.
10
'--~
10- 7
50
MPa • m1/2
Effect of temperature and humidity on fatigue crack growth in 180-mm (7.0-in.) 5083-0 plate.
As shown in the above graph, growth rates for alloy 5083-0 are appreciably higher in moist air than in dry air. Growth rates in water solutions of sodium chloride are similar to those in moist air.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William w. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 195
347
348
12-30. Alloys 5086-H34, 5086-H36, 6061-T6, 7075-T73 and 2024-T3: Comparative Resistance to Axial-Stress Fatigue 0.7 R~O
Frequencv = 1.1 kHz 0.6
0.5
.
U
~
0.4
....
~
.
~
.~
0.3
a:
0.2
0.1
5086-
5086-
6061-
7075-
2024-
H34
H36
T6
T73
T3
Ratio of axial-stress fatigue strength of aluminum alloy sheet in 3% NaCI solution to that in air.
Fatigue strengths of aluminum alloys are lower in corrosive environments such as seawater and other salt waters than they are in air, especially when evaluated by low-stress, long-period tests. As shown in the above bar chart, such corrosive environments produce smaller reductions in fatigue strength in alloys of the more corrosion-resistant types, such as 5xxx and 6xxx alloys, than in less resistant alloys, such as those of the 2xxx and 7xxx series. Like stress-corrosion cracking of aluminum alloys, corrosion fatigue requires the presence of water. In contrast to stress-corrosion cracking, corrosion fatigue is not appreciably affected by test direction, because fracture resulting from this type of attack is predominantly trans granular.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 220
12-31. Alloys 5083-0/5183: Fatigue Life Predictions and Experimental Data Results for Double V-Butt Welds IOOr---r-,--...--I'""T'"T'TTT"--,----.-r....T"T"TT,----r-,--...--I'""T'"TT"n ~ ~
60
400
40
300 200
.: 1II
10 8
CT,=+18kSi
6
KI~
4
5083 -0/5183 Double - V Butt Welds
2
I 4 10
K
s~s
' mol =2.60, 0ls0.0Iln., R =0, 1= 3/8 .n. CT, • + 18 ksi
30 20
10
10~
NT' Cycles
Total fatigue life predictions and experimental results for 5083-0/5183 3/8-in. (10-mm) butt welds.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 114
349
350
12-32. Alloys 5083-0/5183: Predicted Effect of Stress Relief and Stress Ratio on Fatigue Life of Butt Welds 1OO~"""""""""""'--''''''''''''''''''''''~-~''''''''''''''''''''''''''''''"T''T"r--''''''---'---'''''''''''''''''''''''''600 80
eo 40
~3-o1!l183
Bull Weld
Kt..." 3.HI.
400 300
°1" 0.0111\ (0.2~41Ml) ."90",' "SO", '" 112ir\ 112 7mm)
- - .,.5,
200
--- .,·0
R"O
j
~-.
2
S~S 1(,... '
30 20
10
Predicted effect of stress relief and stress ratio on 5083-0/5183 butt weld fatigue life.
Because of the high notch-root plasticity during the first few cycles, before the material cyclically hardens, the aluminum weld considered here (5083/5183) exhibits little dependence upon either residual stress or stress ratio, even though the relaxation of the stabilized mean stress (uos) is very slow-as indicated in the above chart.
Source: F. V. Lawrence, "The Predicted Influence or Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 113
351
12-33. 7XXX Alloys: Cyclic Strain vs Crack Initiation Life 100
.fIN AllOY 7075·T6 7050·T6 7075·T7 7050-T7
W 0 :::l f-
::J Q.. ::iE
UTS kSi (MPa) 810(558) 887(612) 73.2(505) 74.1(511)
<>:
TYS ks; (MPa) 769(530) 82 7(570) 658(454) 63.8(440)
./, fl. 2" 95 12.0 11.0 14.0
z
a:
f-
en ...J
10- 2
/7050.T6
<>:
:-~-~ ----. --"""""''''''''""""",""
f0 f-
-------
"""'''''''''''''1;;./11
1O-3 ':'":-_ _---'100
-'-
-':-_ _----J'-_ _--'-
-'-_ _............
10 7 REVERSALS TO INITIATION, 2N
Cyclic strain versus initiation life for laboratory-fabricated high-strength 7XXX aluminum alloys,
Plots of elastic-strain amplitude versus life have seen relatively little use for commercial alloys, but plots oflog total strain amplitude versus life have been used more frequently to compare materials. This approach offers the advantage that both high- and low-strain fatigue may be characterized with one plot. As illustrated above, fatigue resistance at low total strain amplitude is governed by the elastic-strain amplitude. Fatigue lives for total strain amplitudes less than about 5 X 10-3 generally increase with increasing strength. On the other hand, fatigue lives for total strain amplitudes greater than about 10-2 generally increase with increasing ductility.
Source: T. H. Sanders, Jr., and J. T. Staley, "Review of Fatigue and Fracture Research on High-Strength Aluminum Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, p 472
352
12-34. Alloy 7050: Influence of Alloy Composition and Dispersoid Effect on Mean Calculated Fatigue Life
Zn
Mg
Cu
5.5
2.2
2.3
to
to
to
6%
2.4%
2.4%
Aged 3 h at 121°C (250°F) + 9 h at 163°C (325°F) 1.0
Alloy 7050 sheet
6%Zn 2.2% Mg 0.4% Mn
1.2
Low humidity
Aged 3 h at 121°C (250°F) + 9 h at 163°C (325°F) Alloy 7050
1.0
0.8 High humidity 0.6
0.4
0.2
% Zr %Mn
0.1
0.4
0.1 0.4
%Cu
1.0
2.3
1.0
2.3
Effect of dispersoid type (based on composition) on fatigue crack propagation life of 7050 alloy sheet.
The influence of alloy composition on dispersoid effect is shown in the above bar chart. The general trend in this chart is that for more finely dispersed particles the fatigue crack propagation life is increased. Whereas dispersoid type appears to have a relatively small effect on mean calculated life, the smaller precipitates provided by aging produce a much larger effect. There is some evidence that new processing practices may provide the fine microstructures needed to enhance fatigue resistance. The potential of intermediate working (commonly referred to as ITMT treatments) remains attractive but has not been proven for notched specimens.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, p 192
12-35. Alloy 7050: Effect of Grain Shape on Cycles to Failure
o
Q.
~
.400 UJ
o
::J I...J Q.
«~ 3 (J) (J)
•
UJ
• 7050 AR • 7050 HR
•
II:
:n
200
... 105
CYCLES
106
TO FAILURE
Stress-life curves for two 7050 alloys having fine, equiaxed grains (AR) and pancake-shaped grains
(HR).
As indicated in the above graph, grain shape showed no perceptible difference in life over a range of stress amplitudes.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 238
353
354
12-36. Alloy 7075 (TMP. T6 andT651): Effect of Thermomechanical Processing on Cycles to Failure
300
~ od~ \
m a.. ~
e
tl c
200
E
0\0
&
\
6-
0 "-
\6,
~
'
0 0'0Q,
""4_ 6 100
L..-
10
TS 632 627 573 567
\
'6
\ 0 6'h
CJ)
.~
\
\~ \
If) If)
PS 587 600 516 488
• 7075 TMP ... 7075 TMP o 7075-T651 67075-T6
"""'---
'lIQ .Q..
--_~~ """'---
.........
........
4
Cycles to failure
Effect ofthermomechanical processing (TMP).on the unnotched fatigue properties of the commerical AI-Zn-MgCu alloy 7075. PS = proof stress (MPa); TS = tensile strength.
Detailed studies of the processes of fatigue in metals and alloys have shown that the initiation of cracks normally occurs at the surface. It is here that strain becomes localized due to the presence of pre-existing stress concentrations such as mechanical notches or corrosion pits, coarse (persistent) slip bands in which minute extrusions and intrusions may form, or at relatively soft zones such as the precipitate-free regions adjacent to grain boundaries. Density has also been found to improve the fatigue performance of certain alloys, although this effect arises in part from an increase in tensile properties caused by such a treatment (see above diagram). It should be noted, however, that the promising results mentioned above were obtained for smooth specimens. The improved fatigue behavior has not been sustained for severely notched conditions, and it seems that the resultant stress concentrations override the more subtle microstructural effects that have been described.
Source: I. J. Polmear, Light Alloys. Edward Arnold Ltd. London, England, and American Society for Metals, Metals Park OH, 1981.P 41
12-37. Alloys 7075 and 7475: Effect of Inclusion Density on Cycles to Failure
1:1
400
Q. ~
ILl
0
;:)
I-
300
:J
Q. ~ c:(
I/) I/)
200
ILl II::
l-
I/)
100
104
105
106
CYCLES TO
107
108
FAILURE
Effect of inclusion density on the stress-life behavior oftwo 7XXX alloys: high-inclusion density, alloy 7075; low-inclusion density, alloy 7475,
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 233
355
356
12-38. Alloy 7075: Effect of TMT on Cycles to Failure
n
400
'E z
7075
oY;I~Nm-' ]
7075TMT 600
~
u
:gw
300
a: ~
til
Cl Z
~
z
200
a: w
~
«
100-'----,..---.-----,-----.----.--10' 10' 10' 107 10' CYCLES TO FAILURE
Influence of TMT on S·N curves (R = -1).
There is evidence in the literature that a uniform dislocation density introduced by cold working improves the fatigue life also in connection with FTMT. The above graph shows an example taken from the work of Ostermann. Most of these improvements are due to an increased yield stress.
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, p 195
12-39. Alloys 7075 and 7050: Relative Ranking for Constant Amplitude and Periodic Overload
fi . • .• .• .• .• •
?<
::'::)
2.0 X 106
Total life
c~ns.tant amplitude life Life Increase due to retardation
Overload ratio = 1.8 Applied every 4000 .cycles
,
Overload ratio = 1.8 Applied every 8000 cycles
,
'0
j
1.0 X 106
E ::::J
Z
0.5 X 106
Constant amplitude (Overload ratio = 1.0)
Overload ratio =1.4 Applied every 4000 cycles I
Relative ranking of fatigue life of 7075 and 7050 aluminum alloys under constant amplitude and periodic single overload conditions.
Crack-growth retardation is caused by tension overloading during fatigue testing. The variable-amplitude test is believed to be more sensitive to alloy difference, and it clearly provides more useful information for alloy-development investigations. For example, as illustrated by the data for alloys 7075 and 7050 in the above graph, quite different results are obtained in constant-amplitude tests than in tests with single overloads every 4000 or 8000 cycles. Thus, information on the variation in load level during fatigue cycling is required for correct characterization of the fatigue behavior of aluminum alloys.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for Metals, Metals Park OH, 1982, P 197
357
358
12-40. Alloy 7075: Effect of Environment and Mode of Loading o
~ 160
AIR
• NaCl
'0
;f
-0
---
.0... - - 0 -
.; 120
'"
~ a: I;; u
::;
80
u
~
MODE I LDADING '"
,I
7 10
o
AIR
•
Nael
;f .;
'" '" ....a: '" u
120
uJ
::;
--_~ •
80
u >u 40
0 ....._ 10~
MODE J LOADING
........-1.-1.........................._
o 0
--0
• • ........- - ' - - '.....................' - - _........- - ' - - ' - ' "....., ...............
Fatigue behavior of 7075 aluminum alloy in air and aerated sodium chloride solution: (above) under mode 1 loading; (below) under mode 3 loading.
Tests performed on a commercial 7075 alloy in a mode 3 loading condition (torsion) indicated that the reduction in fatigue resistance associated with cathodic charging was considerably less than it was under mode I loading (note above charts). Although total immunity to corrosion fatigue was not observed, the slight reduction in fatigue resistance can be associated with conditions that did produce a true mode 3 loading condition both on a micro-scale and on a macro-scale. To summarize the aluminum alloy results, it appears that corrosion reactions liberate hydrogen, which effectively embrittles the region in the vicinity of a crack tip. The specific details of the embrittlement are not known, but it appears that dislocation transport of the hydrogen is involved. It has been speculated that hydrogen may collect at the semicoherent precipitate-matrix interface, thus explaining the reported fracture plane; however, a great deal more research will have to be performed before a more definitive answer will be available.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 356
12-41. Alloy 7075-T6: Effects of Corrosion and Pre-Corrosion -------------------
200
A
7075 AI T6 0.5 M NoCI 276 MN/m (40 ksil mean stress RT
- 28
24
N-
E 150
20
<,
z
~
'iii
V> V>
~
w cr 100
l-
V>
u u u>-
:i 50 -
- 16
""
12
~
A AIR B CORROSION C PRE CORRODED/AIR FATIGUE D PRE CORRODED/HEAT TREATED /AIR FATIGUE L.LLL.J...U _ _L I I I l l d _ _ .I_-.l...0--4 5 10 106 10
8
4
0 7
10
The effects of corrosion and pre-corrosion on the fatigue lives of a 7075T6 alloy. Note that re-solutionizing and re-aging the alloy after precorrosion results in a significant increase in fatigue resistance.
Fatigue resistance of high-strength aluminum alloys is severely affected by corrosive solutions, especially chloride solutions, 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. Experiments on a 7075-T6 commercial alloy and on a highpurity analog of the alloy (AI-5.0Zn-2.5Mg-1.5Cu) indicate that localized hydrogen embrittlement may be responsible for the poor corrosion fatigue resistance of these alloys. For example, 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). If specimens are pre-corroded and tested in laboratory air, there is also a significant reduction in fatigue resistance (curve C). The reduction in life at low Nfis associated with pits which form at nonmetallic inclusions. If the alloy is re-solutionized and aged, equivalent to a low-temperature bake, a significant amount offatigue resistance is regained, indicating at least partial reversibility of the damaging phenomenon and strongly suggesting a solid-solution effect arising from environmental interaction.
Source: D. J. Duquette, "Fundamentals of Corrosion Fatigue Behavior of Metals and Alloys," in Hydrogen Embrittlement and Stress Corrosion Cracking, R. Gibala and R. F. Hehernann, Eds., American Society for Metals, Metals Park OH. 1984, P 265
359
12-43. Alloy 7075: Effect of Cathodic Polarization on Fatigue Behavior
2U
.-... ,. -,
~"
' ~~'
I~
.
L .. .' ,.
.....
10
~
. / " " 13v Na2S04
13v NaCI
.---- .
~
N,
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; however, if the alloy is cathodically charged during stressing, sulfate ions prove to be equally damaging, particularly at long NJ , At lower NJ the slight decrease observed in cr solutions appears to be associated with damage to the passive film, as shown in the above S-Ndata, In SO~ solutions, a crack must initiate to break the protective film to allow access to the bulk alloy, Cathodic charging of the high-purity analog of the 7075 alloy also shows a reduction in fatigue resistance, In many cases, fatigue crack initiation in the equiaxed-grain high-purity alloy is intergranular, and at more active cathodic potentials there is a tendency toward a higher percentage of transgranular cracking.
Source: D. J. Duquelle, "Fundamentals of Corrosion Fatigue Behavior of Metals and Alloys," in Hydrogen Embrittlernent and Stress Corrosion Cracking, R. Gibala and R, F. Hehemann, Eds., American Society for Metals, Metals Park OH, 1984, P 266
361
362
12-44. Alloy 7075-T6: Effect of Surface Treatments and Notch Designs on Number of Cycles to Failure
o PAlMlIIC ACIO " ANODIZED AND WATER SEALED
50
41
~
~
40
"j
ij
!:1'"
}I
I;;;
......'------------~ }O
I~
4 10
4 10 NUMBER OF CYCLES
lOB
Tension fatigue test of 7075-T6 aluminum alloy sheet, notch factor K T = 1. ·}D
o PAlMIIIC ACID " ANODIZED AND WATER SEALED
21
~
~20
o
"
ol-
0
0
----------
ij ~15
t;; 10
4
10
I~
104 NUMBER OF CYClES
Tension fatigue test of 7075-T6 aluminum alloy sheet, notch factor K T = 2.37.
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.
The three charts show the effects of notch designs and surface treatments on fatigue properties of aluminum alloy 7075-T6 sheet; the table shows the effects of 17 surface treatments.
Surface treatment
No. of cycles to failure
Polished . 125,000 Anodized and water-sealcd ••• 125,000 Propionic acid •••••••••••••• 2,800,000 Vale ric acid •••••••••••••••• 15,000,000 Caproic acid •••••••••••••••• 9,200,000 Octanoic acid ••••••••••••••• 12,300,000 Decanolc acid .. 7,500,000 Lauric acid ••••••••••••••••• 8,600,000 Myristic acid ••••••••••••••• 11,600,000
Surface treatment
No. of cycles to failure
Palmitic acid ••••••••••••••• 30,000,000 Stearic acid •••••••••••••••• 8,700,000 Docosanoic acid ••••••••••••• 6,000,000 Sebacic acid •••••••••••••••• 13,700,000 Octyl alcohol ••••••••••••••• 6,000,000 Dodecyl alcohol ••••••••••••• 7,000,000 Dodecylamine •••••••••••••••• 18,500,000 Hexandeiamine ••••••••••••••• 3,000,000
Sheet was anodized: 15% sulfuric acid, 23 °C, 15 amp/sq ft, 40 minutes. Stress amplitude: 26,000 psi.
Source: Irvin R. Kramer. "Improvement of Metal Fatigue Lifebya ChemicalSurface Treatment, "in Fatigue-An Interdisciplinary Approach, John J. Burke, Norman L. Reed and Volker Weiss, Eds., Syracuse University Press, Syracuse NY, 1964, pp 250,251
364
12-45. Alloy 7075-T6: Effect of R-Ratio on Fatigue Crack Propagation
1 -
R
I
I
I
o
I
3
10
0.00
.20x
I;
.33e .50l> .700
10- 1
f--
f--
c
a... ~
.80-
10
2
o
R 0.00 .20 • .33e .50l> .700 .80 -
f--
Q)
U
>.
Q)
U >. o ..... E E
.....u 10- 2
E E
f--
~10
z .....
" C
"
lie: ~
It
10- 3 f--
I
...
1-
~ ~
~ ..... .g ~
o o
101
o
_
o
2
5
10
20
30
I
I
I
3
10
30
tJ.K (MPa Ifffi )
AK(MPavm)
Fatigue crack propagation in aluminum alloy 7075-T6 showing the effect of R ratio and the applicability of the Forman, Keraney, and Engle relation. The scatter in the data is much less in the latter.
The above diagrams show that data scatter is much less when the rate daldN is computed according to the equation due to Foreman et al. R< O. The proposed Foreman equation is: da C(t::.K)1I1 dN
(I -R) K c -
s«
Source: Marc Andre Meyers and Krishan Kumar Chawla, "Mechanical Metallurgy: Principles and Applications," Prentice-Hall, Inc., Englewood Cliffs NJ, 1984, p 716
12-46. Alloy 7075: Effect of Predeformation on Fatigue Crack Propagation Rates
VACUUM -2
10
oJ
.9!
10
u >.
~
E E
z Ci "tl "tl
,~
7
10
12
15 20 11K [ MNm-JI2 )
30
40
50
Influence of predeformation by cold rolling on fatigue crack propagation rates for 7075. Ih 100°C, SEN specimens, vacuum, R = O.I,j= 30 Hz.
Cold deformation also increases the fatigue crack propagation rate as shown in the above graph, which compares an undeformed structure with 10% and 20% cold rolled structures.
Source: G. Liitjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds.. American Society for Metals. Metals Park OH, 1980, P 207
365
366
12-47. Alloys 7075 and 2024-T3: Comparative Fatigue Crack Growth Rates for Two Alloys in Varying Humidity 10"
c:--,---,--,r-r-,.,..-----,---r--____:::>
4
78910
a«
20 1
(MPa m / 2 )
Comparative fatigue crack growth rates for aluminum alloys 2024-T3 and 7075-T6 in air ofvarying humidity.
Relationships between rate of growth of fatigue cracks and stress intensity for the alloys 2024-T3 and 7075-T6 are shown above. Other 2xxx series alloys show rates of crack propagation similar to tha t of 2024-T3 over most of the range of test conditions. In general, these alloys have rates of crack growth that are close to one-third those observed in the 7xxx series alloys. It is now common to use precracked specimens to assess comparative resistance of alloys to stress-corrosion cracking, since this type oftest avoids uncertainties associated with crack initiation.
Source: I. J. Polmear, Light Alloys. Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH, 1981,p79
12-48. Alloy 7075-T651: Fatigue Life as Related to Harmonic Generation FATIGUE LIFE EXPENDED
50 o
30*>
60*
1000
2000
I..LI 0
l:::Ji=? 4 -' - '
~~
3
2 00 LL.
I..LI::?: =>"" -'
2
>0 >£2
L58 a.. I..LI Vl
00
FATIGUE CYCLES
Peak value of normalized second harmonic generation as a function of fatigue life.
Recent experiments on flexural-fatigue specimens (aluminum alloy 7075-T651) clearly show the potential of harmonic generation for fatigue monitoring. The above chart shows the peak value of the harmonic generated as a function of fatigue life. At 60% ofthe fatigue life expended, the harmonic had increased by about a factor of four.
Source: O. Buck and G. A. Alers, "New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals, Metal Park OR, 1979, P 137
367
368
12-49. Alloys 7075-T6 and 7475-T73: Effect of Laser-Shock Treatment on Fatigue Properties
.25" OPEN HOLE
0_
.75" 0.0.
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 = .t
LASER-SHOCKED
CONTROL SPECIMENS SPECIMEN NO.
1
NO. OF CYCLES TO FAILURE 51,500
n,loo
2 3
SPECIMEN NO.
NO. OF CYCLES TO FAILURE
1 2
473,800 520,700
AVG. SCATTER
497,250 1.1
385,000
AVG. SCATTER
169,600
7.5
FATIGUE TEST RESULTS FOR 7475-T73 ALUMINUM "MAX = 20 KSI NET R= • I
LASER-SHOC KED
CONTROL SPECIMENS SPECIMEN NO. 1 2 3
AVG. SCATTER
NO. OF CYCLES TO FAILURE 41,500 74,300 109,300 75,033 2.63
SPECIMEN NO.
NO. OF CYCLES TO FAILURE
1 2
171,800 266,200
AVG. SCATTER
218,950 1.5
The fatigue test specimens were 0.25 inch thick by 1.5inches wide and approximately 9.5 inches long, as shown in the above sketch. The specimen blanks were laser-shock processed, and then the 0.25-inch-diameter hole was bored through the center of the laser-shock-processed area. The diameter of the laser-shock-processed area is three times the fastener hole diameter. All of the specimens had machined surfaces of less than 125RMS. All of these open-hole specimens were fatigue tested to failure at a maximum net section stress of20,000 psi, and an R= 0.1 under constant-amplitude load control. Three control specimens for each material were tested to establish the typical fatigue life for the material. Two LSP specimens were tested for each material to establish the degree of improvement due to the laser-shock processing. The fatigue test results for the 7075-T6 material are summarized in the upper tabulation. The LSP specimens showed three times better fatigue lives on the average and much less scatter than the unprocessed material. The results for the 7475-T73 material are summarized in the lower tabulation; these show the same typically large increases in fatigue life and reduced scatter. It should be noted that the 7075-T6 material shows better fatigue resistance than the 7475-T73 material, whether or not it is laser-shock processed. This is largely due to the differences in dislocation/precipitate interactions that result from the T6 and T73 heat treatments. The dislocations appear to shear through the precipitate particles in the T6 condition. The precipitate particles are apparently so strong in the T73 condition that the dislocations just loop around the particles.
Source: William F. Bates, Jr., "Laser Shock Processing of Aluminum Alloys," in Source Book on Applications of the Laser in Metalworking, Dr. Edward A. Metzbower, Ed., American Society for Metals, Metals Park OH, 1981,pp 256-258
12-50. Alloy 7075-T6: Effect of Laser-Shock Treatment on Hi-Lok Joints 3/16~ HI-LOKS & TAPER-LOKS (4 PLACES)
0-
-0
c:::::=:=========$$:Jr====
.072"
Joint Fatigue Test Specimen Geometry 106
.....-- - - - - - - - - - - - - - - --== 0=----. 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. The purpose of this test was to evaluate the fatigue life improvement of laser-shack-processed fastener holes when the holes are loaded by the Hi-Lokfastener in bearing. A secondary purpose was to find out if the cheaper Hi-Lok fastener system in a laser-shack-processed hole would show as good a fatigue life as the much more expensive Taper-Lok fastener system. The above bar chart shows the test results for three different stress levels. At each stress level, three specimens with standard holes and three specimens with laser-shack-processed holes were tested. The specimens tested at the l4-ksi stress level showed severe fretting at the intersection of the hole wall with a badly galled area of the fretted faying surface. All of the fatigue origins occurred at or near the hole wall corners on the faying surface.
Source: William F. Bates, Jr., "Laser Shock Processing of Aluminum Alloys," in Source Book on Applications of the Laser in Metalworking, Dr. Edward A. Metzbower, Ed., American Society for Metals, Metals Park OH, 1981, pp 262-263
369
370
12-51. Alloy 7075 (High Purity): Effect of Iron and Silicon on Cycles to Failure
400
n I
E
z
:::E u If) If)
300
UJ
a:
Iii o Z
200
~ Z
a:
UJ
~
<
100-'--,----r-----.,-----.-----,-10' 10· 10' 10' 10' CYCLES TO FAILURE
Influence of Fe and Si content on SoN curves (R = -1).
The large Fe- and Si-containing inclusions are detrimental to the fatigue life of smooth specimens, because these inclusions serve as easy nucleation sites for cracks. Comparing two alloys, one containing these inclusions (Commercial Purity 7075) and the other one not (High Purity 7075), shows the improvement in fatigue life due to the removal of these inclusions (see the above SoN curves). The alloy termed High Purity 7075 in this figure still contains Cr and therefore the small Cr-containing inclusions. This is important because the removal of these small inclusions would have the opposite effect on fatigue life.
Source: G. LUtjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds.• American Society for Metals, Metals Park OH, 1980, p 193
12-52. Alloy X-7075: Effect of Grain Size on Cycles to Failure o 0..
::r
300
UJ
a
...::> ::::; 0..
::r « Vl Vl
200
UJ
a::
Vi 100L~,----------~;----------'-=--------~;-
o
__
CYCLES TO FAILURE
Influence of grain size on S-N curves (R = -1, f= 100 Hz) for X-7075 with PFZ (20 hat 160°C, or 320 OF).
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. Again, the tensile yield stress was equal for both grain sizes. 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.
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals. Metals Park OH. 1980, p 192
371
372
12-53. Alloy X-7075: Effect of Grain Size on Stress-Life Behavior
400
" c,
::E
w
300
0
~
::::i
a, ::E
lI) lI)
w
0:
200
l-
ll)
CYCLES TO FAILURE
Aluminum alloy X-7075; 24 h at 100 °C (212 OF).
The above chart shows the grain-size effect in a stresscontrolled test for a high-purity 7075 alloy (X-7075) aged to contain shearable precipitates. Since the flow stress is determined by the interaction of dislocations with the coherent precipitates, the yield stress is approximately the same for both alloys. Opti- . cal examinations of the specimen surfaces show that cracks nucleate much earlier in specimens with the large grain size. Cracks nucleated at intense slip bands for both grain sizes.
Source: Edgar A. Starke. Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure." in Fatigue and Microstructure. American Society for Metals. Metals Park OH, 1979. P 225
12-54. Alloy X-7075: Effect of Environment; Air vs Vacuum
-I
10
..
A: 24h 100'C
VACWM
C:48h 180'C
AIR
n
~
u
E
E u
-)
10 _,
10
Z
:E a '0
-i
10
I06..L..----,--r--..,.....--..--..,.....---.------r----r--..,...... Influence of environment (laboratory air) on fatigue crack propagation rates for underaged (A) and overaged (C) condition. X-7075, CT specimens,R = O.I,f= 30 Hz.
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. To illustrate this point, the above graph shows the comparison between underaged and overaged microstructure also for tests performed in laboratory air. The aggressive environment has a much more pronounced effect on the underaged condition, leading even to an opposite ranking of the alloy conditions. In laboratory air the cracks propagate still along slip bands at low dajdN rates.
Source: G. LOtjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 204
373
374
12-55. Alloy X-7075: Effect of Environment on Two Different Grain Sizes
-2
10
~
u
GS 200.lJrn GS
-J
46.lJrn
10
>-
.!2
E
-4
E 10
z
~
.g
-5
,,/~
n
3.5NoCI
10
I06.L----.-----.---'-r---..-----r-----.---...----.7
10
15
20
30
40
50
t:. K [MNrn-3I2 ) Influence of environment (3.5% NaCI) on fatigue crack propagation rates for two different grain sizes. X-7075, 24 h 100 DC, CT specimens,R = 0.1,1= 30 Hz.
The same tendency is observed for the grain size dependence of crack propagation ifthe tests are carried out in a 3.5% NaCI solution (note above curves). The influence of environment is larger for the large grain size. For this highly aggressive environment the cracks propagate at low dal dN rates along grain boundaries in a complete brittle fashion.
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 204
12-56. Alloy X-7075: Effect of Grain-Boundary Ledges on Cycles to Failure ..-
c
Q.
!
LLJ
o
300 r--~"-T'"""...-r'1""""'-r---'----r-"'T"""T"T"l"T~---'----"--"-"""""""'"
co..
••
\
'PO \
~ 250 I---~ID-----':~_----_--t-----____l
:J Q.
~ ~
\
b
~o. ....
"
200t------O-~~--""'~-+__---____l
X-7075 150 t----::-----±-:-~-:-----t-----_l 0---0 If = 0, 8h 160°C --If = 0.5,4h 160°C 105 CYCLES Effect of grain-boundary ledges on the stress-life behavior of an alloy containing nonshearable precipitates and PFZ.
One method that may be employed to reduce the slip length in the PFZ is thermomechanical processing. If enough cold deformation is employed to introduce steps (or "ledges') into the grain boundaries, 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. The above chart shows the results of a stress-controlled test for two high-purity 7075 alloys, one cold-worked 50% to produce grain-boundary steps. The cold work drastically reduced the incidence of grain-boundary cracking and improved the fatigue life at high stress amplitudes. At low stress amplitudes and long fatigue lives, crack nucleation occurred at inclusions for both alloys. This effect is most likely due to stress concentration at inclusions.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 230
375
376
12-57. Alloys X-7075 and 7075: Effects of Chromium Inclusions on Fatigue Crack Propagation
24 h 100·C
-1
VACUUM
10 n
~ u
'"
-)
10
~
E E u
z
-.
10
~
a
"0
-5
10 -6
10
7
40
50
Influence of Cr-containing inclusions on fatigue crack propagation rates by comparing aluminum alloys X-7075 and 7075.24 h 100 °C, CTspecimens, vacuum,R = 0.1,! = 30Hz.
As shown above, 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. Furthermore, they normally increase fatigue crack propagation rates also indirectly by their effect on grain size and shape.
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 207
12-58. Alloy 7475-T6: S-N Diagram for a Superplastic Fine-Grain Alloy 80
r------------------------------. o SPF Conditions:
= As pr......d A = SPF - highsup.rpl.stic st"inl o = SPF - low luporpl.stic st"inl
T= 516·C i=21104S-1
70
-+ = No,.ilu"
R =+0.1
60
AS PROCESSED S·N CURVE
MAX STRESS 50 (Ksi) 40
o
AOC>AO
o
_
-...
6.
0
...
(]I
_-------0-+0-+
3D
20 I-.....I._..I....I..I-I._..................I-.....I._...."""'"-"u.._"'-..........I."""".........._ 1~ ,~ 10' 10' 10'
....."""'"-"U 10'
CYCLES TO FAILURE
S·N curves for a superplastic aluminum alloy: fine-grain 7475. All testing was done with smooth specimens.
Tests on fine-grain 7475 alloy have shown improved fatigue life as superplastic strain is increased, as shown in the above SoN diagram. An even more dramatic improvement is obtained in damage tolerance.
Source: C. Bampton, F. McQuilkin and G. Stacher, "Superplastic Forming Applications to Bomber Aircraft ." in Superplastic Forming. Suphal P. Agrawal. Ed.. American Society for Metals. Metals Park OH. 1985. p 77
377
378
12-59. Alloy 7475: Effect of Alignment of Grain Boundaries on Cycles to Failure 300 0
,..... 0 Q..
~
-..J
250
I
- II
UJ
0 :J
-<9=0.5
!
~
...J Q.
o <9=0
200
16h 160°C 6h 160°C
~ <{
I/l I/l UJ
0:
0
150
~
II)
0
7475
o I
CYCLES
1.
I
I
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.
If the stress axis is parallel or perpendicular to the long grain dimension, there will be no shear stress parallel to the grain boundary, and preferential deformation within the PFZ will be restricted. 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.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 232
12-60. Alloy 7475-T6: Superplastic vs Nonsuperplastic, as Related to Fatigue Crack Growth 1 X10-] r - - - - - - - - - , . - - r - - - - - - - - ,
7475 T-6 FINE GRAIN ~ 1 X 10-' (.)
~
.....
7075 T-6 -
(J)
w
::I: (.)
Z
1 X 10- 5
Z
o
<,
< o
1 X 10- 6
1 X 10- 1
1.----I._.L.I.....L..J..J.._...l----JL...-..JL.I...l.----I._.....L....J...J..J
1
10
100
1000
K (I-R)M-l (KSI "'lINCH)
This comparison of conventional, coarse-grain, nonsuperplastic aluminum alloy 7075 with superplastic alloy 7475 shows almost an order-of-magnitude reduction in crack growth for the superplastic material.
Source: C. Bampton, F. McQuilkin and G. Stacher, "Superplastic. Forming Applications to Bomber Aircraft," in Superplastic Forming, Suphal P. Agrawal, Ed., American Society for Metals, Metal Park OH, 1985, p 77
379
380
12-61. Alloys X-7075 and 7075: Effect of Chromium-Containing Inclusions on Cycles to Failure
400
:g
300
~
I-
:::; Q.
~
~
200
~ J~----,--;,------------'--r:---""""'----""'-3
10
I
I
10'
10
5
I
la'
•
CYCLES TO FAILURE
Influence of Cr-containing inclusions on SoN curves (R = -1,1= 100 Hz) comparing aluminum alloys x-7075 and 7075, 24 h 100 0 C. (Arrows indicate crack nucleation visible by LM at u a= ± 200 MNm -2. j
The above S- N curves compare results obtained from testing commercial 7075 alloy with the alloy X-7075 which does not contain Cr. These small inclusions, as in the tensile test, inhibit the formation of intense slip bands, thus retarding crack nucleation as indicated by arrows on the graph. Due to these small Cr-containing inclusions, the grain size of the 7075 alloy was somewhat smaller as compared to that of X-7075, which also may have contributed to the observed improved fatigue behavior.
Source: G. LUtjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 195
12-62. Aluminum Forging Alloys: Stress Amplitude vs Reversals to Failure
"t
E
IJ. o • x
400
z
2.:
~
7075-T73 squeeze formed SF1-T61T73 squeeze formed 6082-T6 squeeze formed 6082-T6 extruded bar
SF1-T6/T73
§300~-T6 S ~
~200
.~~o x
~
~
In 10 3
.
10 4 10s 10 6 REVERSALS TO FAILURE (2N,)
SON fatigue data for several squeeze-formed forging-type aluminum alloys compared with extruded AA 6082-T6.
The above chart presents results from push-pull, about mean zero, fatigue tests which have been carried out on a servohydraulically controJled machine. The tests have been carried out on samples cut from actual components, not from separately made testpieces. The results from conventionaJly extruded AA 6082 (H30) are included for reference: in this case, the data are in the longitudinal direction, it not being possible to obtain samples of sufficient size from the transverse direction. The results show good fatigue properties for squeeze-formed material, which in one case compare favorably with conventionally extruded material. This further substantiates the claim that squeeze formings in general are comparable with forgings with respect to mechanical performance.
Source: G. Williams and K. M. Fisher, "Squeeze Forming of Aluminium-Alloy Components," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, p 367
381
382
12-63. AI-5Mg-O.5Ag: Effect of Condition on Fatigue Characteristics 180 ~------'--------r------.,----, 0.2% proof Tensile AI-5%Mg-0.5%Ag stress IMPa) strength IMPa)
150
ST and quenched
•
85
Aged 1 day 175°C
200
310
175
270
X Aged 70 days 175°C
lU
a..
~
(/] 120 (/] Q)
~
(/]
Cl
c
'16
0
90
c
•
~ .~ oho ~ ~.... 0_
~
~
L60
10&
87 X 72
7 10
48
~ lOll
Number of cycles Fatigue (SoN) curves for the alloy AI-SMg-O.SAg in different conditions.
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. It is well known that binary AI-Mg alloys such as AI-5Mg, in which the magnesium is present in solid solution, display a relatively high level of fatigue strength. The same applies for an AI-5Mg-0.5Ag alloy in the as-quenched condition, and the above diagram shows that the endurance limit after lOS cycles is ±87 MPa, which approximately equals the 0.2% proof stress. This result is attributed to the interaction of magnesium atoms with dislocations, which minimizes formation of coarse slip bands during fatigue. The silver-containing alloy responds to age hardening at elevated temperatures due to the formation of a finely dispersed precipitate, and the 0.2% proof stress may be raised to 200 MPa after aging for one day at 175°C (350 OF).
Source: I. J. Polrnear, Light Alloys. Edward Arnold Ltd. London, England, and American Society for Metals, Metals Park OH. 1981, P 42
12-64. AI-Zn-Mg and AI-Zn-Mg-Zr: Effect of Grain Size on Strain-Life Behavior 100
r----------------..., a,
" ", GI .':::.
.. _- " ........
24h at 150GC ...........
::---'.::: .... ' ...
(\J
-...
e--
...... ,
:.:~~
Q.
......
<.&>
• Failure AI-Zn-Mg -Zr , 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. The AI-Zn-Mg alloy had large grain size; the AIZn-Mg-Zr, small grain size.
The above chart shows Coffin- Manson life plots of two averaged AI-Zn-Mg alloys. The small-grained AI-Zn-Mg-Zr alloy has a much longer life than does the large-grained AI-Zn-Mg alloy. The improvement in life is attributed to increasing the cycles to crack initiation, as indicated in the chart. A convergence is noted for long lives (low plastic-strain amplitudes) for this strain-controlled test. Since the fine-grained material hardens more than the other at low strains, the stress to enforce the applied strain is greater at long lives, and this affects the life improvement due to the fine grains.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979. P 228
383
384
12-65. AI-Zn-Mg: Strain-Life Curves of a Large-Grained Alloy
• 4h at 120°C .96h at 150°C
10. c..
wiN
..........
AI-Zn-Mg
0.1
10 1
Strain-life curves of large-grained AI-Zn-Mg alloy having shearable precipitates when underaged (4 h at 120 ° e, or 250 OF) and nonshearable precipitates plus PFZ when overaged (96 h at 150 "C, or 300 OF).
Since the strain localization occurs in a region free of solute, overaging the matrix precipitates or adding dispersoids does not homogenize the deformation. This is clearly illustrated by comparing the Coffin Manson life curves of underaged and overaged specimens of largegrained AI-Zn-Mg alloy (see above chart). The tensile yield strength and strain to fracture are approximately the same for both specimens. The underaged alloy has shearable precipitates, which results in strain localization, the formation of intense slip bands, and early crack nucleation under cyclic loading. Overaging was one method described for homogenizing deformation; however, this method is not effective for large-grained material. Preferential deformation in the PFZ also leads to strain localization and results, for this particular case, in the same fatigue life. Dispersoids distributed throughout the matrix would not inhibit strain localization in the PFZ for the same reason.
Source: Edgar A. Starke, Jr., and Gerd Liltjering, "Cyclic Plastic Deformation and Microstructure,"in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 227
12-66. Aluminum With a Copper Overlay: Stress Amplitude vs Cycles to Failure
-
70
g Q. 60 ~
D
-b
....... 0
<,
50 +1
c-
<,
~o
<,
o Al D
<,
_-- -----
Al (Cu)
0
<, 0.....
40 10
4
10
5
6
10
10
7
Stress amplitude U a versus number of cycles to failureNj for AI and AI with a Cu layer. Note the pronounced improvement in the latter at large N j •
The fraction of fatigue life spent in crack nucleation, N; / NJ , increases with decreasing load amplitude (i.e., at high N's). It would be expected that the treatment suggested above will produce a great effect in large fatigue life regimes (i.e., under conditions where initiation of fatigue crack is more important than its propagation). The above graph shows this phenomenon in the case of pure aluminum and aluminum with a copper surface layer.
Source: Marc Andre Meyers and Krishan Kumar Chawla. "Mechanical Metallurgy: Principles and Applications," Prentice-Hall, Inc .. Englewood Cliffs NJ, 1984, P 707
385
386
12-67. P/M Alloys 7090 and 7091 vs Extruded 2024 60
400 l>
50 'in
"'"
40
~
30
..
20
E OJ E
'x ::;;;
o
7091·T7E69 (2 lots) 7090·T7E71 Open symbols-longitudinal Solid symbols-long transverse
60°
O~;" 300
- - I ndicates did not fail /to:D
10
.
o,
::;;;
Notch tip radius = 0.013 in. KT = 3.0 4
~-
~
200 E OJ E
..
'x ::;;;
Band for 2024·T351 rod and bar
100
Stress ratio R " 0 Ambient air
Ol-_ _-'-_ _--'--_ _--'-_ _-'-_ _--'-_ _--'-_ _--'_ _--.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.
The notched axial fatigue strengths of alloys 7090 and 7091 are 35 to 40% higher than those of alloys 7050, 7075 and 2024 (an 11M alloy often selected for its resistance to fatigue) at one million or more cycles.
Source: Robert H. Graham, "Wrought Aluminum PIM Alloys," in Powder Metallurgy-Applications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 240
12-68. P1M Alloys 7090 and 7091 vs 11M 7050 and 7075 Products 60
r-----.-----.----.-------.----.-------.---~-------,400
A 7091-T7E69 (2 lots) 7090-T7E71 Open symbols-longitudinal Solid symbols-long transverse
o 50
]
40
E
30
60°
O~" 300 ..
Indicates did not fail Notch tip radius KT = 3.0
g
=
a.
0.013 in.
2:
~~ 200
:>
~
E :>
E
E
..
'x
20
10
2:
Band for 7050 and 7075 products
Stress ratio R = 0 Ambient air'
100
O'----:----'-,-------'-c-----'-:c-----L,---'--:---'----:----'-,------' 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.
The notched axial fatigue strengths of alloys 7090 and 7091 are 35 to 40% higher than those of alloys 7050,7075 and 2024 (an 11M alloy often selected for its resistance to fatigue) at one million or more cycles, as shown above.
Source: Robert H. Graham. "Wrought Aluminum PIM Alloys." in Powder Metallurgy-Applications. Advantages and Limitations. Erhard Klar, Ed.• American Society for Metals, Metals Park OH, 1983. p 240
387
388
12-69. P1M Aluminum Alloys: Typical Fatigue Behavior
9·71B-in. R o o
Forged
o
x
'u;
40
Forged 601AB-T6
?:
£-=fL-i-----3-"1!I
201 AB-T6
0.300-in. diam
c.
201AB-T6
202AB_T2~
E
.~ 20
x 2''"
60 1AB-T2'>.
601A~
_ _..::::===_=====a=_=_=_=_=_-1
OL------''------''-----'----'-----'-----'----'
10'
10'
10· Cycles to failure, N
10'
Typical fatigue behavior of alloys 601AB, 201AB and 202AB.
Fatigue is an important design consideration for P / M parts subject to dynamic stresses. The above S-N diagram shows typical fatigue behavior of specimens of alloys 6OIAB, 20lAB and 202AB in the T2 (as-cold-formed after sintering) and/ or T6 tempers.
Source: John D. Generous and Wayne C. Montgomery, "Aluminum PI M-Properties and Applications." in Powder MetallurgyApplications. Advantages and Limitations, Erhard Klar, Ed.. American Society for Metals, Metals Park OH, 1983, p 214
12-70. P1M Aluminum Alloys: Comparison With Specimens Made by Ingot Metallurgy 500
~ 1.9;mm (0.075
:
E--t--]- E= f-- t tf--+-1=
7.62 mm . _ _ 8.39 mm (0.300 in.) (0.3.30 ln.I diarn. diarn.
400
'"
::E
~
•
..
300
1;;
E
::>
-1.0
in.)
60
12.19 mm (0.480 in.)
I
I
Notch tip radius: e ~ 1.254 mm (0.01 in.) K, ~ 3
-,
0..
Ii
70
60· sharp' V
251 mm (9'10 in.)
50
40
0
.'"
]
~ 1;;
E
::>
E
'x
E
'x
0
'" ::E
'" ::E Smooth
200
30
0
a---
100
..
Solid lines represent bands for 7075·17352 !three lots), longitudinal direction Dashed lines represent bands for 2014-T61 (nine lots smooth. five lots notch I. longitudinal direction
0 10'
-- -
10'
10'
•
---
10'
20
10
-~
10'
0 10'
Cycles
Rotating-beam fatigue strength for die forgings of P 1M alloy X709I-T7E76 and ingot metallurgy alloys 7075-T7352 and 2014-T6I. For P/M X709I-T7E76: 0 ,smooth, transverse direction; • ,notched, transverse direction; - denotes test specimen did not fail in number of cycles indicated.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society ForMetals, Metals Park OH, 1984,p 469
389
390
12-71. P 1M Aluminum Alloys: Comparison With Forged 7175 for Cycles to Failure .500
t."
L10."" J,~.L
mm
10.,)'
~t~
400
-
70
in.]
60
Notch tip radius:
e = 0.33 mm (0.013 in.)
-
=3
K,
ul
50
Ul
e 1il
300
-
E E
:>
'x co ::E
-" ul
Ul
e 1il
'iii
40
E
'x co
• m:a 0
200
•
-.we
7175-T73~ ~
•
•
•It
•
30
o • • •~
~
10'
r r 103
10'
::E
20
Band for 100 _forgings (six lots) Stress ratio: R = 0.0
o
E
:>
10
o 10'
10'
10'
10'
10'
Cycles
Comparison of axial-stress notch fatigue strength of P 1Malloy X7091-T7E69 die forgings and ingot metallurgy alloy 7175-T736 die forgings. 0 , longitudinal direction, one lot; • , short transverse direction, two lots; - denotes test specimen did not fail in number of cycles indicated. Stress ratio: R = 0.1.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,P 469
12-72. Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack Growth Rates; Air vs Salt Water
'(f3_-------------------"'" crack orientation L - T specimens: SEN , GNP, DCB , CT R • 0.0 - 0.1 , 0.1 - 30 Hz ambient temperature
11 AI - alloys in salt water: 2048-T851 2219 - T 87 2618 -T6 5456-H117 5456-H321 6061 -T651
~ 2cro - T651 2024 - T 3 2024-T~
2048 - T851 2219 - T851 2219 - T87 20 cyclic stress
18 AI- alloys in air:
2618 - T 6 5456 -H117 5456-H321 6061 - T651 7'005 - T63 7039 - T6X31 30
7005 - T63 7050 -T736 7fJ15 - T651 7175 - T736 7475 -T651
40
7050 - T 736 7075 - T 6 7079-T651 7106 - T 63 7175 - T 736 7475 - T 651 50
60
iltensity range, 6K, [MN 'm-¥2]
Comparison of scatterbands of corrosion-fatigue crack growth rates and fatigue crack growth rates of many commercial aluminum alloys.
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 615
391
392
12-73. Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack Growth Rates in Salt Water 163~------------------,
7475 - T651 7005 -T63 2618 - T6
7175 - T 736 7075 - T 651 2219 - T 87 5456 - H 321
,----, ~
~ ~
10- 5
E
'------'
~~
6061 -T651 2048 - T851
-6
10
7050 - T736 5456 - H 117
10- 7
Q)
~ typical experimental scatter
::J
.21 ....
....ro
crack orientation L - T specimens: CNP I DCB I CT environment: salt water ambient temperature R ·0,0-0,1 I 0,1 - 30 Hz -10 10 0,......--+---+----+----+----+---..... 10 20 30 40 50 60 cyclic stress intensity range,
~K , [MN. m-~]
Corrosion-fatigue crack growth rates in salt water for aluminum alloys exceed the scatterband.
As shown in the above graph, curves for growth rate are somewhat higher than the air-test scatterband, but at very low and very high stress-intensity ranges, no significant difference between fatigue and corrosion fatigue crack growth rates is observed.
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material, "in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, p 614
12-74. Various Aluminum Alloys: Wrought vs Cast, and Influence of Casting Method on Fatigue Life
500 400 300 250
'" ~
e
200
<;;
150
rf.
~
t:
en
~ ~
201---"': 18t----
~
161----+~
§
141----+----,.:-'
E
~ iU
":;'l
1.
~
II)
100
121----'--+---'--~
75
1 0 1 - - - - + - - ' -......- +-..........:
8!---'---'-f--'-----if--"""'""-,..-t----'
71-----+--.-:.+------1------+--,,-:~
50
61-----+---+--...--:'":-...;,......1----'---+..,.....-..:.,......+---~
106
lOS
109
Life N (cycles (log))
Representative S-N curves for various aluminum alloys are shown in the above graph. Note the absence ofa sharply defined "knee" and true endurance limit. This is typical of nonferrous metals. In the absence of an endurance limit, the fatigue strength at 108 or 5 X 108 cycles is often used. (To give a "feel" for the time required to accumulate this many cycles, an automobile would typically travel nearly 400,000 miles before anyone of its cylinders fired 5 X 108 times.) As is true for most metals and alloys, the wrought versions of aluminum alloys have greater fatigue strength than the cast (see graph). 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).
Source: Robert C. Juvinall, Fundamentals of Machine Component Design, John Wiley & Sons, New York NY, 1983, p 207
393
394
12-75. Aluminum Casting Alloy AL-195: Interrelationship of Fatigue Properties With Degree of Porosity
26
... 24
... ...
, , 1\
,~
" 4
16, [,>..
22
,,1,- 3,
,
... ...
5) ~ 1-1',
~
S-N CURVES CASTING ALLOY AL - 195 DEGREE FATIGUE* DEGREE FATIGUE* POROSITY STRENGTH POROSITY erRENGTH - - 5 - - 9,000 ;t 2 10,500 6 8,150
rr;ooo
2
1'>,
,""1\
i"
,~
I,
" ...
~
x 18
"
f' ~"
41- I - -
- .
' ... , 8./
...
~
...
~I'
....
"of-.
~
f'
'"
I"'-\l:i: I'
"o
.......
8 } DEGREE :-:-- POROSITY
10 Q
,,5 10
. I-
~
",I'. ,
1'1· 10
I
-
"I",
~
I""' l ......
"o
12
A~~~
-- f - -
.~
....
...
-- -
8
,~
r-,
_~
"'I'
1,000
1
.... REPRESENTATION OF NO FAILURE
'I\.
<,
,'\..
~ .......
6
9,000 9,000
* AT 10 000 000 CYCLES
\,
,1"\
20
3 4
1)
IV
,,"'.
;.. "~ r-; ... <, ~
..
I-
...
r-,
r>-
J'.... ..... ~ ..... I-
.. .. .. r-...... .....-r--
10,,6
-
101
Interrelationship offatigue properties with degree of porosity for AL-195 casting alloy.
Source: N. E. Promisel, "Evaluation of Non-ferrous Materials," in Materials Evaluation in Relation to Component Behavior (Proceedings of the Third Sagamore Ordnance Materials Research Conference), Syracuse University Research Institute, Syracuse NY, 1956, P 65
12-76. Aluminum Casting Alloy LM25-T6: Squeeze Formed vs Chill Cast; Effect on Reversals to Failure 400
N
I
E
z
~300 w o
=>
l-
ii 200 2
« (/)
~ 100
LM 25-T6 chill cast
a::
In O'-:--_--L 2
10
1~
--'4
m
'-::-_ _--'-
1~
--'---'
1~
REVERSALS TO FAILURE(2N,) S-N curves for aluminum casting alloy LM25-T6; chill cast versus squeeze formed.
Fatigue tests have been carried out with LM25 samples, which were cut from a bracket component. A servohydraulically controlled fatigue machine was used to execute push-pull tests about mean zero. The results are presented in the above chart, which includes for reference the results of similar tests carried out on conventionally cast LM25. It can be seen that a significant improvement in the fatigue performance has been achieved by squeeze forming this type of alloy.
Source: G. Williams and K. M. Fisher, "Squeeze Forming of Aluminium-Alloy Components," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, p 367
395
396
13-1. Copper: Effect of Air and Water Vapor on Cycles to Failure 5.6
<,
s: u
c ...5.4
Vi Q;
~5.2
'" c
--
Ol C <0
a:
"E
4.6
OJ (J)
~ -, ~
o
OJ
I
---
I
l'
'~
I--
::f 5.0 e Vi '0 4 .8
~"
<,
\
-
......
Q(~I'("
1110
~fiedAir ....
_______ T-D
r-... I' ~
----
If) 0 .... PlJr i f · iltr)fJ
I
I---
II
10
- --
~-- 1---
~
'ea"
--I" "~~. InAir"_
4.4 6
10
-
.... iA
~
~
\.\~I \
-
~
7
---
-
~ 5.10
7
Endurance, Cycles to Fracture (Log. Scale] - - + Indicates Specimen Unbroken
The effect of air and water vapor on the fatigue life of annealed copper.
The effect of atmospheric oxygen on fatigue life of copper has been investigated; oxygen and water vapor reduce fatigue life in copper. 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). Based on these experiments, the investigators concluded that: Fatigue cracks form early, because the majority of life is concerned with crack propagation (environment has little or no effect on nucleation and initial growth). 2 Oxygen and water vapor are the primary damaging constituents in air (water vapor alone is effective). 3 Oxygen must be a gas (preoxidation or intermittent exposure is not effective).
Source: D. J. Duquette. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 336
13-2. Copper: Applied Plastic-Strain Amplitude vs Fatigue Life
Cu
• prestrain 20"1. • annealed "prestrain 40/.
_-
A-
Applied plastic-strain amplitude versus fatiguelife curves for copper at long life.
Helgeland was the first to observe and claim a fatigue limit for copper (actually the plateau stress, although it was not recognized as such at the time). Unfortunately, his results were apparently contradicted by those of Kettunen, who observed failures at stresses down to 17.7 MPa. This difficulty was resolved by Laird, who showed that Lukas and Klesnil's long-life CoffinManson plots showed failures to occur only down to the plasticstrain fatigue limit; at lower strains, no failures were observed in the testing time available (see above chart). However, Lukas et al. also carried out stress-cycling tests, in which they monitored the plastic strain. Specimens that had been stress-cycled yielded a plot of saturation plastic-strain amplitude versus life, where failures occurred at strains as low as 10-5 . The difference between these tests is that in strain cycling, the stress is low in the initial cycles and increases to saturation, whereas in stress cycling, full application of the load in the first cycle causes a large strain in a soft material. This initial large strain creates the PSB cell structures, which would not otherwise form in a constant-strain test. Since Kettunen applied the full load to his specimens, failures were observed at stresses below that of the plateau. Helgeland, on the other hand, although he was stress cycling, imposed a low stress at the start of his tests and increased it gradually to the chosen value.
Source: Campbell Laird. "Mechanisms and Theories of Fatigue," in Fatigue and Microstructure, American Society for Metals. Metals Park OH, 1979, P 195
397
398
13-3. Copper Alloy C11 000 (ETP Wire): Effect of Temperature on Fatigue Strength 200 r - - - - - - - - - - - - - - - , - - - - - - - - - - - - - - - , 26
'"
Cl.
::E
22 ]1
150
s:."
...en
s:."
'C,
e
e
e
18
.
1;;
Q)
::J
::J
en
en
'+>
'"
u.
~ '+>
100
14
'"
u.
10
Stress cycles
Rotating-beam fatigue strength of electrolytic tough pitch copper, CllOOO wire, 2 mm (0.08 in.) diam, H80 temper when tested at various temperatures.
Source: Metals Handbook, 9th Edition. Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 289 .
13-4. Copper Alloy C26000 (Cartridge Brass): Influence of Grain Size and Cold Work on Cycles to Failure 60
50
'"
<,
a.
0 0 0
40
-
I'--~
....0 ::l
><
II!,,,)" +~
20
..... 1'0- ...
.........
r-,
Q)
l J)
"0
<,
........
.... 30
<,
I'
~
~~~
...e
~ a
t;.1--
~?" I"'--.:!""...
1'; I-
Q)
--
f_
:: ..
r-rx_
IJ..
-i
x
-+
+
10
6
10 107 Cycles for Failure ,N Legend o Group G (60% Cold I:>. Group D (40% Cold o Group A (20% Cold • Anneal I Grain Size ... Anneal 3 Grain Size x Anneal 4 Grain Size + Anneal 5 Grain Size
Drown) Drawn) Drawn) 0.0/2 mm. 0.026 mm. 0.051 mm. 0.131 mm.
Influence of grain size and cold work on fatigue strength of copper alloy C26000 (cartridge brass).
Changes in grain size and in degree of cold work which result in increased tensile strength or hardness usually result in improved fatigue strength. The above S-N curves illustrate this and indicate the influence of grain size and cold work on the fatigue strength of alpha brass. All specimens were prepared from the same heat and, therefore, had the same nominal composition.
Source: George M. Sinclair, "Some Metallurgical Aspects of Fatigue.vin Fatigue-An Interdisciplinary Approach, JohnJ. Burke, Norman L. Reed and Volker Weiss, Eds., Syracuse University Press, Syracuse NY, 1964, P 69
399
400
13-5. Copper Alloy C83600 (Leaded Red Brass): S-N Curves; Scatter Band 180 ~------..---------r--------''-------~_-----~
25 160 f---~~~-+-------+------+-------+--------l
rf
140 f------~
20 ]
:::E
~ ~
.J::.'
t»
s
1;;
1201--------+------'
1;;
Ql
:::l
Ql
'"
:::l
.~
'"
.~
u.
100 f - - - - - - - + - - - - - - - + -
------+-------1
80 f-------+--------+-------I-----=
0-
15
10
Stress cycles
S-N curves (scatter band) for copper alloy C83600 (leaded red brass).
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 406
u.
13-6. Copper Alloy C86500 (Manganese Bronze): S-N Curves; Scatter Band 400
If 300 :E
~c e
t:
200
Cll
;:)
.~
'" u.
100
- 50
- 40
~~ ..................
--
~
.r."
",
c
30
~~ n
e
t:
Cll
- 20 .;::;c: u. - 10 '" ;:)
Number of stress cycles S-N curves (scatter band) for copper alloy C86500 (manganese bronze). All testing was performed at room temperature.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, ArnericanSociety for Metals, Metals Park OH, 1979, p 35
401
402
13-7. Copper Alloys C87500 and C87800 (Silicon Brasses): S-N Curves; Scatter Band
.
300 40
D-
~
.,;
Ii
~
1;; Cl
..
c: ';;
E
... Ql
<{
';;; ~
250 30
200
.....~
Cl
.
e
';;
E Ql
150 100 104
20 105
106
107
108
~
109
Stress cycles S-N curves (scatter band) for copper alloys C87500 and C87800 (silicon brasses) tested at room temperature.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 416
13-8. Copper Alloy C92200 (Navy "M" Bronze): S-N Curves; Scatter Band
'"
200
0-
25
::i:
Ii 150
~
s
20
'"
Ci:
...~..
Cl
Cl
...'"E
'iii .>t
15
100 50 104
105
106
107
108
sc ..... '"
10 Ci:'" 109
Number of stress cycles
S-N curves (scatter band) for copper alloy C92200 (Navy "M" Bronze, or steam bronze) tested at room temperature.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, p 421
403
404
13-9. Copper Alloy C93700 (High-Leaded Tin Bronze): S-N Curves; Scatter Band 180
25
.
0~
't!'
e
160 140
20
Cl
.
.g
E l!l
120 ~_----=l15
100 80 104
105 Number of stress cycles
S-N curves (scatter band) for copper alloy C93700 (highleaded tin bronze) tested at room temperature.
Source: Metals Handbook•. 9th Edition. Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH. 1979, P 426
405
13-10. Copper Alloy No. 192: Effect of Salt Spray on Tubes liAs Received" Cycles to Failure
Tube Oia. 3/16-io. 1/4-io.
5/16-io.
Cycles to Failure After 180-0ays Exposure to Salt Sorcv 3/8-io.
Tube Die, 3/16-io. 1/4-io.
5/16-io.
3/8-io.
I
I----+I---J----o 0
-r----
i
-~ I :0
0 0
0
0 0
0
10
10
0
~
b
0
'Q,
-
0
a 0 0
..!
uc-,
U
.....
~
M 0
-
N
~
Results of fatigue tests on copper alloy tubes before and after salt spray exposure.
The tubes made from the copper alloy failed in the range of 105 to 106 cycles. After exposure for 180 days to salt spray, the fatigue performance level was not lowered (see plot above). Brazed steel tubes, prior to salt exposure, failed in the same test in the range of 105 to 107 cycles. After 30-days exposure to salt spray, the resistance to fatigue was 105 to 106 cycles. After 90-days exposure, the steel tubes showed no fatigue strength in this particular test.
Source: Donald K. Miner, "An Effective Solution to the Problem of Hydraulic Brake Line Corrosion," in Source Book on Copper and Copper Alloys, American Society for Metals. Metals Park OH, 1979, p 356
406
13-11. Copper Alloy 955: Goodman-Type Diagram
1.5 UTS: 152025 ksi
1----+----+----+-----t----2F------i
150
120 UTS: 101.5 ksi ~
0
°ili °ili
90
iii
.....
""
~
60
30 I-:JIC---I
..,.<........,h~---fI+_-I+_---1t+_-H__-I::l=-_H-+++-_I_If--I
~
o
°i
.3
G _----' 30 '--_
o
30
---'-
-'-
60
90
-L.
120
--'-_ _- - - '
150
180
Goodman-type diagram (after Creech) for annealed copper alloy 955.
Two variable-speed, flat-plate testing machines of the fixeddeflection type were used for the test work. These machines have a speed range of 750-2000 cpm with a maximum deflection of I in. The yield, ultimate tensile strength and 1.5times the ultimate are plotted in the graph above. The fatigue limit at zero mean stress was determined and was found to be 22.0 ksi.
Source: J. M. Cieslewicz. "A Modified Goodman Diagram to Predict the Fatigue Limits of Copper Alloy 955." in Source Book on Copper and Copper Alloys. American Society for Metals. Metals Park OH, 1979, p 40
14-1. Magnesium Casting Alloy QE22A-T6: Effects of Notches and Testing Temperature 150
If. :2
125
+1
100
t:
~
75
E ::J E
50
..
'x :2
-. \.
~
20
<,
1
/Unnotched
~ +1
"",-...............
U·nolched /(K t ~ 2)
1
~. E E
::J
V·notched /IKt = 31
'x ~
25
Cycles of stress
150
If.
125
:2 +1
~.
100 75
E ::J
E
'x ~
50
- 20
<,
<;
<,......
-
I
/20 oC (68°FI- 16 ~
I
.,
i! 200°C 1392°F)_ 12 :;; E ~25~ °c (482 of) ::J E 'x ~
25
Cycles of stress
S-N curves for magnesium alloy sand castings, QE22AT6, showing effects of type of notches (upper graph) and testing temperature (lower graph). Rotating-beam (Wohler) tests. Machine speed was 2960 Hz.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 589
407
408
14-2. Magnesium Casting Alloy QH21A-T6: S-N Curves; Effects of Notches and Testing Temperature 25
175 150
rr.:::;;
125
...
~
100
1;;
E ::l E
.
'K :::;;
\
20 °c (68 IF)
-, r-,
20
............... '-- Unnolched "-
<,
75
I'-...
50
] 15
--
U·nolched K t=2110
...
g
E ::l E 'K
.
:::;;
25
106 Cycles of stress
175
I
25
Unnolch.d
150
rr.:::;;
+~ 100
~
~ E
~
\
20
125
75
'I\.
"'---
-,
]
20°C (68 OF) 15
~50 °c 1480 OF)
I"--.
10
01
g E ::l E 'K
.
:::;;
50 25
Cycles of stress
S-N curves for magnesium alloy castings, QH21A-T6, showing effects of notches (upper graph) and testing temperature (lower graph). Rotating-beam (Wohler) tests; machine speed 2960 Hz.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 590
409
14-3. Mg-AI-Zn Casting Alloys: Effects of Surface Conditions on Fatigue Properties Meanstress. ksi
Meanstress,ksi
20 ±20
±20
.16
.16 ~
Ii
:t12
-l---I----l
e ~ ."
~
.
!i
4-----11-----l
t---+--+-----j---t---+--....=j±4
100
!II
~
ell
f---t---+---+----'=-+--+--, '4 125
125 Mean stress,MPa
M&an stress,MPa Meanstress,ksl
Mean stress, ksi
o
10 15 20 .150,------,,--,-----,,---y-,-----...---,---r-.
20 10 '150 r ---,----i---.---'T...---.:n--r--T--. O
±20
±20
Cast plates '1251---+---+-- 107 cycles. loaded in bending .16
"6
Xi
~-
.12
tl2
---f----f----;
--+--1---;
I---t---+----f---""'---+-...:....-=J '4
!II
1-----J-=...."""-+-----1F"'-...;...f---+--..., ±4
125 Mean stress,MPa Mean stress,MPa
Effect of type of surface on fatigue properties of cast Mg-AI-Zn alloys.
Machining improves fatigue properties of castings, as shown in the above curves. Small radii, notches or fretting corrosion are more likely to reduce fatigue life than are minor variations in composition or heat treatment.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 532
'" .~
.§ ~
410
15-1. Molybdenum: Fatigue Limit Ratio vs Temperature 0.9
0.8
~
0.7
~r\
!
0
';:;
a::'"
0.6
\
0.5
0.4
o
100
200
300
400
500
600
Temperature,OC Ratio of the fatigue limit of molybdenum at various temperatures to its tensile strength at the same temperature.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society for Metals, Metals Park OH, 1979, P 774
16-1. Tin-Lead Soldering Alloy: S-N Data for Soldered Joints 5000
I
4000
I'~
......
~~ A
'iii
3000
--....r- ....
a. ui
'" ~
in
2000
~
.......
:--.... ........
~ :--....
~ ~
"'r-..... ~
-- r- --...........,
,
I~
<,
1
2
2
4 6 8
......
f4.. ......
'Speed of testing I .... ~/min I I 2
. . 'N .........
~~
1000
a
.....
2
102 Cycles to failure
..... ...........
4 6 8
10 3
':).04 0.2
-
2
The fatigue strength of soldered joints is a complex and difficult subject to examine. Because solder alloys are strain-rate sensitive and have large elongation capabilities, the performance of fatigue tests under constant stress causes progressive and rapid relaxation of the joint, and conversely, tests under constant strain do not reflect a practical application situation. The influence of the rate of stress cycling in terms of rate of straining on the fatigue life of copper soldered joints with 60%Sn-40%Pb alloy is presented in the above graph.
Source: Metals Handbook, 9th Edition. Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park OH, 1983, P 1095
411
412
16-2. Babbitt: Variation of Bearing Life With Babbitt Thickness Babbitt thickness, in.
0.030
0.040
o Bimetal I
• Trimetal
~ OJ
:::l
...... ..
200
Cl
',J
OJ
.~
Q;
100
a:
0.25
0.50
0.75
1.00
Babbitt thickness, mm
Variation of bearing life with babbitt thickness for lead or tin babbitt bearings. Bearing load was 14 MPa (2000 psi) for all tests.
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. Use is made of this principle (see graph above), in both two-layer and three-layer constructions, in which the surface layer is composed of a lead or tin alloy, usually no more than 0.13 mm (0.005 in.) thick.
Source: Metals Handbook, 9th Edition, Volume 3. Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society [or Metals, Metals Park OH, 1980, P 806
16-3. SAE 12 Bearing Alloy: Effect of Temperature on Fatigue Life Bearing temperature, ° F
300
~
en .;;
200
300
r---,c-r.----~-----,
200 I------'\-t---+------t
.:?
'"
> .;;
'" 100
Gi
I----~r_--t-----f
II:
OL-_ _L--_ _L - - _ - - - - l 200 100 150 50 Bearing temperature,
°c
Varlanon of bearing life with temperature for SAE12 bimetal bearings.
The alloy lining was 0.05 to 0.13 mm (0.002 to 0.005 in.) thick, on steel backing. Bearing load: 14 MPa (2000 psi). As indicated, operating temperature markedly influences fatigue life.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 813
413
414
17-1. Unalloyed Titanium, Grade 3: S-N Curves for Annealed vs Cold Rolled 500
"'
400
a.. :2
'" 300 ~'"
'" E
6,
- 70
r-;
.~~
- 60 Cold rolled
0
.-. .... -
-
~
Annealed
J
.§ 200 )(
"' :2 100
.;;;
50
.><
::i
40 ~
'" E
J 30 E
'x
20 :2"'
-
10
Number of stress cycles
SON curve for unalloyed grade 3 titanium. Data were obtained by rotating-beam testing at room temperature.
Source: Metals Handbook, 9th Edition. Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 376
17-2. Unalloyed Titanium, Grade 4: S-N Curves for Three Testing Ternperatu res 600 500
'"
Q.
~
400 '" '"
~
E ::>
.Sx
'"
~
300
-------
- 80 _40°C (-40 ° F)
-
- 60 20°C (6SoF)
- 50
.;;; ..><
:i
~
E ::>
I 315°C (600 of)
- 40 .Sx 30
200
100
70
'"
~
20
104 Number of stress cycles
S-N curves for unalloyed titanium, grade 4, at subzero, room, and elevated temperatures. Data were obtained by rotating-beam testing of unnotched, polished specimens machined from annealed bar stock.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH. 1980.0378
415
416
17-3. Ti-24V and Ti-32V: Stress Amplitude vs Cycles to Failure
-
STRESS
CI
a..
::lE
CONTROLLED
600
ILl 0
::>
I-
:J a..
Ti-24%V, A.Q.
::lE
400 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- 24%V) and cyclic softening (Ti-32%V).
Cyclic-response curves indicate that the Ti-24%V alloy undergoes extensive cyclic hardening, whereas Ti-32%V undergoes cyclic softening, as indicated above. Hardening is caused by incomplete reversibility of twinning.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society for Metals, Metals Park OH. 1979, p 236
17-4. Ti-5AI-2.5Sn: Effects of Notches and Types of Surface Finish
...
800
a::2: 600 ~
~l>
~o
t: 400
rt:-
:I
e~
60 t:
-
Ground
Hand finished
E
.>I.
80
~
~
e
100 .;;;
Shot peened
E
40
.§ 200 )(
E
';(
...
20 ~
::2:
o
8 10 Lifetime, stress cycles
...
a::2:
l:/"
e t:
800
Ultrasonic machined
100
Slab milled
600
80 Chem milled and annealed Chem mille
400
E E 200 ';(
60 40
:I
::2:
0 104
...
105
..
106
107
800 100 ]
....... 600
e
t: 400 E
:I
... )(
::2:
20 ::2:... 0 108
Lifetime, stress cycles
a::2:
,§
E :I E
';(
...
,:;
... ...... ~
.>I.
200
~are
- 80
notch (K r = 2.4)
~~
~
a
l>Al>A
A
V-notch (K r = 3.2) 'I
16 ~
~l>l>
60
-
40
E E
:I
'x - 20 ::2:... 108
o
Lifetime, stress cycles
SON curves for annealed titanium alloy Ti-5AI-2.5Sn (rotating-beam tests). Top and center graphs show fatigue strength for different types of surface finish. Bottom graph shows fatigue strength as affected by type of notch.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Melals, American Society for Metals, Metals Park OH, 1980, p 382
417
418
17-5. Ti-SAI-2.SSn and Ti-6AI-4V: Fatigue Crack Growth Rates Stress intensity factor range, /iK, ksi Viii. 5
10 10-4
Ti-6AI-4V
NI24 to -269 0C - - ---f c:l 10-31---- (75 to - 452 of) ----=",...j~t;f--+_--__l 0C(75 0F) ~ {24 E ELI -196°C (-32°F) E -269°C (-452 of) ~'
~
i
~
~
e
NI ELI
24 to -269°C (75 to -452 of) b 10-41-----+--~~--I~----4-----=1 Cl
~
u
OJ
., :J
.Cl
'" u.
Stress intensity factor range, s«, MPa
.,;m
Fatigue crack growth rates for Ti-5AI-2.5Sn and Ti-6AI-4V.
Data on fatigue crack growth rates for Ti-5AI-2.5Sn and Ti6AI-4V alloys are plotted above. These data indicate that the exposure temperature has no effect on the fatigue crack growth rates for Ti-5AI-2.5Sn and Ti-6AI-4V(NI). However, over part of the 11K range, the fatigue crack growth rates for Ti-6Al4V(ELI) are higher at cryogenic temperatures than at room temperature at the same 11Kvalues.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 765
17-6. Ti-6AI-6V-2Sn: Effects of Machining and Grinding Ti-bAI-bV-lSn (STA, 42 Rcl Surface R aug h ne s s , AA
SURFACE GRIND + PEEN
Ah ...
b5
43 44 70
lO
20
END MILL· PERIPHERAL CUT
Gentle Abusive
150
J67 67
14 14
73
28 39
12
II 145
I 45
I", Off- Sta nda rd
147
'".
85 I
20
48 120
~5
iOff-S'.n""."
o
43 55
)83
,v.
Gentle Abusive
ECM + PEEN
I
Gentle
HAND GRIND
ECM
I
Gentle Cony. Abusive
SURFACE GRIND
I
40
I
I
60
80
I
ENDURANCE LIMIT, KSI
Bar chart presentation showing effects of various machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-6V-2Sn.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 355
419
420
17-7. Ti-6AI-6V-2Sn (HIP): S-N Curves for Titanium Alloy Powder Consolidated by HIP
iii 120 ANNEALED PLATE (MINI
" vi en w
a:
l-
ll)
10('
:2 :> :2 X <{
:2
80
60 -'--
---,
"T"
"T'"
--,
5
10 NUM8ER OF CYCLES
OJ
HIP RUN 1. AS MACHINED (SPEC W "M" SU8SCRIPT WAS EVALUATED 8Y METALLOGRAPHY)
o
HIP RUN 2. VAC ANN AT 1300·FI24 HR
mHIP RUN 2. VAC ANN AT 1300·FI2 HR rn HIP RUN 2. VAC ANN AT 1300·FI16 HR iii
HIP RUN 4. VAC ANN AT 1300"FI24 HR
S-N curves showing endurance limits for titanium alloy powder consolidated by HIP at 900°C (1650 OF).
Note that most data points obtained in this phase fell within the representative data band for annealed forgings. In the specimen designated with an "M"subscript, low fatigue endurance was apparently associated with failure initiation at an inclusion, This shows that a clean powder is required for parts that are fatigue-critical and must operate with the equivalent of fully forged properties.
Source: R. H. Witt and W. T. Highberger, "Experience With Net-Shape Processes for Titanium Alloys," in Production to Near Net Shape: Source Book. C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals. Metals Park OH. 1982. P 277
17-8. Ti-6AI-6V-2Sn (HIP): S-N Curves for Annealed Plate vs HIP
ANN PLATf'
. / eeo F-16KeI-2 /lR
120
\
tOO
tgTReeS, KSI
/
\
80 €JO
\
<,
22-5'0~-fOl<'Sl-I/-IR
'7------:
RAfIIG[ O~ PROGRAM l?ATA
40 20 ~4
~~
~~
~7
CVClP8 TO {"AILURf" S-N curves showing that HIP Ti-6AI-6V-2Sn is equivalent to annealed plate of the same composition.
Source: W. Theodore Highberger. "Manufacture of Titanium Components by Hot Isostatic Pressing," in Production to Near Net Shape: Source Book. C. J. Van Tyne and B. Avitzur, Eds.. American Society for Metals. Metals Park OH, 1983, P 304
421
422
17-9. Ti-6AI-2Sn-4Zr-2Mo: Bar Chart Presentation on Effects of Machining and Grinding Ti-bAl-lSn-4/'.r-lMo (STA, 36 Re! Surface Roughness, AA
SU RFACE GRIND
END MILL-
39 41
-, 68
I-l'G~e.:.:n~tl~e_....,.."....
Cony I 17 ~IO
I
lZO
I-l'G~en~ltI.:i-e
"T""
PERIPHERAL CUT ... A"'lb~u~.,:.:·v~e
o
'"
8Z
36 77
...1 47
I ZO
I
40
I 60
I
I
80
ENDURANCE LIMIT, KSI
Bar chart presentation showing the effects of specific machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-2Sn-4Zr-2Mo.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys,"in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 355
17-10. Ti-6AI-2Sn-4Zr-2Mo: Constant-Life Fatigue Diagram Minimumstress. ksi -100
-50
0
+50
100
150
1200
150
1000
.;;;
800
If.
.><
~ t: E :> E
:;; 10D ~
t:
600
E E
:>
'x
'x
~ 400
~ 50 10 7 cvcles lifetime
200
0 -800
-600
--400
-200
+200
400
600
800
1000
0 1200
Minimumstress, MPa
Constant-life fatigue diagram for duplex annealed Ti-6AI-2Sn-4Zr-2Mo sheet, 1 mm (0.04 ln.) thick.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 385
423
424
17-11. Ti-6AI-2Sn-4Zr-6Mo: Low-Cycle Axial Fatigue Curves
1600
'"
Q.
:2
220
---
1400
STA
200
~.
~
'" E ::J E
1200
'" :2
1000
'x
180 160 140
800 1
103
'"
.>< ~.
~
E ::J E
'x
'"
:2
120 105
Number of cycles Low-cycle axial fatigue curves for Ti-6AI-2Sn-4Zr-6Mo. STA (solution treated and aged) condition: 1 hat 870 ° C (1600 OF), water quench, age 8 h at 595°C (1100 OF) and air cool. DA (duplex annealed) condition: 15 min at 870°C, air cool, then 8h at 540 ° C (1000 ° F) and air cool. All fatigue tests conducted at a stress ratio of R = 0.1. Open symbols indicate fatigue tests; solid symbols, tension tests.
Source: Metals Handbook, 9th Edition, Volume 3. Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals. American Society for Metals, Metals Park OH, 1980, P 395
17-12. Ti-8Mo-2Fe-3AI: S-N Curves; Solution Treated and Aged Condition 600
I
500
'"
e,
:::E
400
'" '" ~ '" 300 E :> E
'x
:::E'"
200
-
80 Tension-tension tests , R = 0,25; K r = 3.5
'.
'in
~ •
60
~
'" '"
~
I'----
.~
- 40
E :> E
'x
,....
'" :::E - 20
100
Stress cycles
SoN curve for Ti-8Mo-2Fe-3AI titanium alloy in the solution treated and aged condition. Data are for 1.S-mm (0.060-in.) thick sheet solution treated 10 min at 790 0 C (14S0 0 F), air cooled, and aged 8 h at 480 0 C (900 OF).
Source: Metals Handbook. 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 403
425
426
17-13. Ti-10V-2Fe-3AI: S-N Curves; Notched vs Unnotched Specimens in Axial Fatigue 1100
I
1000 ~o
~
'" 900
Q.
:!: ~
...e
800
'"
E :::>
.§
700
)(
'" 600 :!:
.---- --
Unnotched
~
~
lili
140
r---
~
'u;
.:,t.
120 A
100
<,
",'
'" ~ '" E :::> E
'x
~ to-..
500
'" :!:
----
80 I--
Stress cycles
500
'"
Q.
400
\~
:!:
~
o RT
~2050C (400°F) '" 100 :!: o 425 °c (800 of)
40 ~
'u; .:,t.
r-,
~
E 200 :::> E
60
Kr = 3.0
\0
'" '" 300
'x
I
Notched
'" '"
~
E :::> E
0
20 'x
'" :!:
I Stress cycles
Axial fatigue of Ti-l0V-2Fe-3AI bar stock in the STOA (solution treated and overaged) condition. Specimens were taken from round bars 7Smm (3 in.) in diameter that had been solution treated 1 h at 760° C (1400 OF), furnace cooled, overaged 8 h at 565°C (1050 OF) and air cooled. Tests were conducted at a stress ratio ofR= 0.1 and a frequency of 20 Hz. Top: results of unnotched bars tested at room temperature. Bottom: fatigue characteristics of notched specimens tested at elevated temperature.
Source: Metals Handbook, 9th Edition.Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 399
17-14. Ti-10V-2Fe-3AI and Ti-6AI-4V: Comparison of Fatigue Crack Growth Rates do/d~ in/cycle (pm/cycle)
10-4-r-----r---...."7T""-------, (25)
MA Ti'-6A/-4V
iI
i i
10-5 (025)
Ti"-IOV2Fe-3A/
i
•
70.6
(0025)/i'... O--.....L.L---
20
(II)
LlK, ksi
(22)
vm.
J, ...J
O'---4..J.O - - - - - - - - '
(33) (44)
(MPoW;;)
Comparison of fatigue crack growth rates. Data for Ti10V-2Fe-3AI,R = 0.05,F= 1 to 30 Hz; for MA Ti-6AI4V,R= 0.08,F= 1 to 25 Hz; for RA Ti-6AI-4V,R= 0.08, F= 6Hz.
Fatigue crack growth rates in air have been found to lie in the scatter band for mill annealed (MA) Ti-6AI-4V, as shown above. At high ~Kvalues, Ti-IOV-2Fe-3Al approaches the performance of Ti-6AI-4V in the recrystallized annealed (RA) condition.
Source: Wayne A. Reinsch and Harry W. Rosenberg, "Three Recent Developments in Titanium Alloys,"in Titanium and Titanium Alloys: Source Book, Matthew Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 375
427
428
17-15. Ti-1 OV-2Fe-3AI: S-N Curve; Notched Bar Fatigue Life for a Series of Forgings Compared With Ti-6AI-4V Plate rrmox,IOJpsi(MPo) 70-r-----~----------____..,
(48J)
60 (414)
50 (345)
40
STA Ti·6AI-4V plate
(216)
30
(201)/i.l-O-4----------'-----"':::OO"------.J
Cycles tofaIlure Comparison of notched fatigue lives for Ti-lOV-2Fe-3AI forgings and Ti-6AI-4V plate. Data for Ti-lOV-2Fe-3AI,R = O.OS,F =KT= 2.9; for STA Ti-6AI-4V plate,R = O.l,K T= 3.
Fatigue characteristics of Ti-lOV-2Fe-3Al are equal to or superior to those of Ti-6AI-4V. Notched fatigue results are shown above. Data from a series of die forgings have shown that the mean value for fracture toughness is 49.1 ksi ~. (54 MPa Jill), with a standard deviation of 2.3 ksi~. (2.5 MPa Jill). K/scc in 3.5% NaCl is typically about 90% of the x;
Source: Wayne A. Reinsch and Harry W. Rosenberg, "Three Recent Developments in Titanium Alloys,"in Titanium and Titanium Alloys: Source Book, Matthew Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 375
429
17-16. Ti-13V-11 Cr-3AI: Constant-Life Fatigue Diagrams Minimum stress, MPa
-100
-80
-60
-40
-20
+20
40 60 80 100 120 140 160 180 200 1400 r-_,_----,~-_,_-_,___,...---.-_r-_r""__,-__,-~""__.--t-er____,"'_,,,._-r____,, 200 180 160 140 120
100
E
"~" 600 ~
80
~
~~
:;; E
~
"K
~
At room temperature
400
60 40
200 - - Unnotched
-
-600
-400
-200
+200
400
-
600
Edgenotched, Kr = 3.0 800
1000
20
1200
Minimum stress,ksi
Minimum stress,ksl 1400
-100
-80
-60
-40
-20
+20
40
60
80
100
120
140
160
180 160 140 ~
0;;;
:;; • 800
120 ""
~~
~E ~
"K
100 600
80
~
E
"E
o~
:;;
60
400 Al315
-c (600 OF) 40
200
- - Unnotched - - Edgenotched, Kr=3.0 -400
-200
+200
400
600
800
1000
20
1200
Minimum stress, MPa
Constant-life fatigue diagrams for Ti-13V-llCr-3AI, STA (solution treated and aged) condition, longitudinal orientation. Data arefor axial fatigue of edge-polished sheet specimens of material solution treated and aged to room-temperature tensile strength of 1203 MPa (174.5 ksi), Corresponding yield strength was 1080 MPa (156.7 ksi); at 315°C (600 OF), the tensile strength was 1078 MPa (156.3 ksi) and the yield strength was 876 MPa (127.0 ksi), Tests were conducted at a speed of 60 Hz.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels. Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, P 401
430
17-17. Ti-6AI-4V: Effect of Condition and Notches on Fatigue Characteristics 800 co
e,
::E 600
::i
~on 400 E :::l E 'j( 200 co
.a...a..
~
-
~ Smooth
--...:
--
bar, STA stock -r:"---~ Smooth bar, annealedrock Notched bar (K t = 3.5). STA stock
"" Notched bar. annealed stock
100 ';;; 80
g
60
on
~
40
::E
106 Number of cycles S-N curves for titanium alloy Ti-6AI-4V (rotating beam) showing effects of STA (solution treated and aged) versus annealed conditions, and effect of notches.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980, p 389
17-18. Ti-6AI-4V: Effect of Direction on Endurance
700 650
m a.. ~ CIl CIl
Q)
;:, CIl
0)
c:
550
.~
'"c: L.
2:! 500 ~ 450
106
Endurance (cycles) Rotating-cantilever fatigue (S-N) curves for three testing directions in 57 mm thick, forged and annealed Ti-6AI-4V bar.
These curves show that fatigue properties are lowest in the long transverse direction. This result has been attributed to the fact that Poisson's ratios are also sensitive to crystal orientation, these ratios being higher in the longitudinal and short transverse directions because stressing occurs parallel to the basal planes. Higher ratios imply greater constraint, which means that the levels of strain will be reduced and the fatigue strength enhanced in these two directions. The differences observed in fatigue strengths in the longitudinal and short transverse directions have been attributed to relative changes in grain shapes that also occur during processing.
Source: I. J. Polmear, Light Alloys. Edward Arnold Ltd, London, England, and American Society for Metals. Metals Park OH, 1981, p 193
431
432
17-19. Ti-6AI-4V: Effect of Isothermally Rolled vs Extruded Material on Cycles to Failure 100 , - - - - - - - - - - - - - - - - - - - - - - - - - -
80
o!';.
iii
'" 60 .,;
w '" ....a:
'" ::; :::l
~ x « ::;
40
AXIAL FATIGUE. RoO. I ROOM TEMPERATURE K,' 2.8 FLAT SPECIMEN WITH 0.050·IN. HOLE
o
20
6.
ISR TEE EXTRUSION
OL-_ _'--_---JL...--l----JL.......I_ _---L_ _---L_-L.........-..L _ _......._ _......._ ...L.............L _ _-'-_ _...... _'--..L-J 103
105 CYCLES TO FAILURE
S-Nfatigue data for isothermally rolled tees versus extruded material. The notched fatigue behavior of the ISR tees is as good as or slightly better than that of the extrusion.
Source: W. T. (Ted) Highberger, Govind R. Chanani and Gregory V. Scarich, "Advanced Titanium Metallic Materials and Processes for Application to Naval Aircraft Structures," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, P 124
17-20. Ti-6AI-4V: Comparison of Wrought vs Isostatically Pressed Material for Cycles to Failure 1200
I 1111
co
o, ~
w'"".""~ grade Ti-6AI-4V ~
1000
I ~ """ ~ ~~ ~ ~ ~ ~
'"
~
U;
~ 125
~
800
~~ ~
~ ~~ ~ ~"
E :J
"~ X
co
~
150
~~ ~
600
~
I I
• Engine mount supports
400 10'
o Witness blocks 1lJ1I Radius failures
~~
~
~ ~i'-
~ ~ t:S ~"
100
~
10'
E :J E
"xco
~
:\I'ci ~ :\" ~~ ~ ~t'- ~ ~ ~ ~"
105
10'
'"
1il
75
10'
Cycles to failure 900
co 700
n, ~
e'"
1il 500 E
I. I I I I o Engme mount supports
Wrought standard grade Ti-6AI-4V
:J
E
"xco
~ 300
100 10'
~ ~~
~ ~" ~ C'\
125
~
~ ~ ~ ~ -<: ~
100
e'"
~
75
~ r-;~ -:~" ~ ~ ~~ ~ C'\~~ ~ ~ ~ ~ ~\ ~ ~ ~ ~ ~ ~~ ~ ~~ r~ ~ 88 ~ ~~
':':";lo,;
10'
"iii
-""
~
105
10'
1il E :J
E
50
~ r-,~ ~"
'x co ~
25
10'
Cycles to failure
S-N curves for titanium alloy engine mount supports. Top: Data are for the standard wrought grade;R = O.I,K,= 1.0, load controlled smooth specimens. Bottom: Data are for isostatically pressed alloy powder, notched specimens;R = 0.1, K,= 3.
Source: Metals Handbook. 9th Edition, Volume 7. Powder Metallurgy, American Society for Metals. Metals Park OH. 1984.P 654
433
434
17-21. Ti-6AI-4V: Effect of Fretting and Temperature on Cycles to Failure
~ z ~
560
w
a:
In 420
Fatigue (all test
~
,om,,,":,,,,
•
R.T
•
200°C
o 340°C
o
z ~280
z
a:
w ~ 140
°1LO"':'~
---..L.::~-----~------.L:;------' 1(f CYCLES
10~
TO
10~
FAILURE
Effect of simultaneous fretting and fatigue ofshot-peened Ti-6AI-4V at room temperature, 200, and 340 °C at a mean tensile stress of 140 MNm·2•
Room temperature fretting was found to have little effect on the fatigue strength at 107 cycles. Fretting at 200 °C lowered the fatigue strength by approximately IS%; furthermore, the fretting fatigue life in the overstress region (70 MNm- 2 above the nonfretted run out stress) was lowered by two orders of magnitude compared with results in the absence of fretting. At 340 °C, fretting similarly reduced specimen life 7 at overstress levels; however, more importantly, the fretting fatigue strength at 10 cycles was reduced to approximately 40% of that found under room temperature conditions. The gross result of fretting normally is fatigue failure brought about by surface damage in conjunction with normal or transient high stresses in a component. It should be said at the outset that visual assessment of fretting mildness or severity is inconclusive by itself, in that the presence of more or less fretting debris on a microscopic examination is not necessarily relevant to the loss of surface integrity. It may in fact be misleading and should not be relied upon for assessing the severity offatigue life degradation. It is the stress state acting in concert with stress raisers (e.g., pits, tears, cracks) which determines the actual fatigue propensity.
Source: Practical Observations of Fretting Fatigue Cracks. p 180
17-22. Ti-6AI-4V (Beta Rolled): Effect of Finishing Operations on Cycles to Failure FATIGUE CHARACTERISTICS OF BETA ROLLED TITANIUM 6AI-4V,
az R c
METAL REMOVAL PROCESSES: SURFACE GRINDING. PERIPHERAL END MILLING. CHEMICAL MILLING MODE: CANTILEVER BENDING. ZERO MEAN STRESS TEMPERATURE: 75' F
U;
80
:<:
en
lol '" 0: f.<
'"
60
~ ~I
-..;
I
I
I I I III
~
............. .... :040
f.< ...l
"-
..;
zo
10 5
III
I
""'-- I
\J
Z E=: ..; Z 0: lol
II
ENDUR. LIMIT
SURF. FINISH
GENTLE MILL GENTLE GRIND
66 6l
41 35
CHEM. MILL
51
zo
ABUSIVE MILL
az
59
ABUSIVE GRIND
13
65
CONDITION
r-, .....
----
10,6
I
I I
II
107
CYCLES TO FAILURE
S-N curves for beta-rolled titanium alloy Ti-6AI-4V. Curves show the effects of the various finishing operations on fatigue.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982,p 354
435
436
17-23. Ti-6AI-4V: Effect of Yield Strength on Stress-Life Behavior 700 Ti - 6AI-4V, 24h 500°C
-.. tJ a..
:::E
650
•
0"0.2 (J"02
910 MPo o eooppm 02 990 MPo • 1900ppm 02
ILl 0
::>
I...J
600
•
Q..
:::E
«
550
UI UI ILl
a:
I-
500
UI
450 4 10 CYCLES
TO FAILURE
Effect of yield strength on the stress-life behavior of two Ti-6AI4Valloys.
In order to establish microstructural effects on fatigue behavior, comparisons should be made on materials having the same yield stress, especially for stress-controlled tests. This is illustrated above where it is shown that two titanium samples having different yield strengths have different stress-life behavior when tested at 500°C (930 OF).
Source: Edgar A. Starke, Jr., and Gerd LOtjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure, American Society For Metals, Metals Park OH, 1979, P 237
17-24. Ti-6AI-4V: Effect of Stress Relief on Cycles to Failure 60
II II
0-AS RECEIVED I e- H T 400"r. 2 HRS.
\
.~
"'" 40
1'.,
iii
II:
o
~ I
\.~ ~
30
Ul
~
-, ~
~
Iii 20
0
kr
~
-
T
- BlOKE IN GRIPS
CYCLES
Flexural fatigue tests of titanium sheet (123,000 T.S.).
Flexural tests of the sheet specimen were made at 1725 CPM. Results are indicated by the open points in the above SoN diagram. The endurance limit was not reached at stresses as low as 20,000 psi.lfan estimated limit of 19,000 psi is chosen, the endurance ratio would be only 0.155, a value considerably lower than for any other known metal or alloy. Most investigators have obtained normal values around 0.3 in similar tests. Several of the sheet fatigue specimens developed fatigue cracks away from the milled specimen edges. The cracks did not appear to be associated with any visible surface imperfection. For these reasons, it was assumed that the sample was abnormal, rather than the test procedure. Very careful oil-powder and fluorescent powder tests, supplemented by metallographic examination, failed to reveal any surface cracks, even when the sheet was flexed to open any incipient hairline defects. It was considered possible, though not probable, that residual stresses from cold rolling were acting in a deleterious manner. If so, a moderate temperature stress relief might help. Brief experiments soon disclosed that temperatures at least as high as 400 OF(205°C) did not lower the hardness; in fact, the hardness may have increased very slightly. Knowing this, a set of sheet fatigue specimens was stress relieved for two hours at 400 OF(205°C). The solid points in the graph above represent the results obtained with these specimens. The endurance limit was not altered significantly. A definite shift to the left in the upper portion of the curve was evident, although the direction of shift was opposite to that, had the heat treatment released undesirable stresses.
Source: Titanium Symposium, Office of Naval Research, p 97
437
438
17-25. Ti-6AI-4V: Interrelationship of Machining Practice and Cutting Fluids on Cycles to Failure lee.
E 7EUI
~--
. e __
~ _
---==='~II --
~~ e.
.Iee
.28e
--_. _--
..
C ~_==
-- •••••
-
--.~--
----
....
-
--..-----... --= ----=
.60e I.ee 2.e8 CYCLES TO FAILURE (IN MILLIONS)
le.ee
6.00
Alternating stress vs cycles to failure in high cycle fatigue of machined titanium surfaces using neutral, chlorinated, and sulfurized soluble cutting fluids. A = abusive grinding; B = low stress grinding; C = end milling.
Influence of Chlorinated and Sulfurized Cutting Fluids On High Cycle Fatigue Properties of Ti-6Al-4V Machined Surfaces at 75°F
Sol. Cutting Fluids
10 7 Cycle Fatigue Strength (ksi) Low Stress Abusive End Milling Grinding Grinding
Neutral
75
62.5
12.5
Chlorinated
65
57.5
12.5
67.5
12.5
Sulfurized
Source: V. A. Tipnis and J. D. Christopher, "Machinability Testing for Industry," in Machinability Testing and Utilization of Machining Data, American Society for Metals, Metals Park OH, 1979, p 26
17-26. Ti-6AI-4V: Relative Effects of Machining and Grinding Operations on Endurance Limit Tio 6AI-4V B ETA ROLLED, l2 R c Surfac e Roughness. AA
SURFACE GRIND
I 62
Gentle ~13
HAND GRIND
r;~ntl~
END MILLEND CUT
Gentle
80 80
157
I 30
Ab .. etve
164
r;~ntl~
I 66
Ah ... lu~
~""n",
CHEMICAL MILLING
nu_~·
o
67 84
177
u ........
END MILLPERIPHERAL CUT
35 65
'II 59
I l2
...
20 165
51
14 5 I
I
I
I
20
40
60
80
I
ENDURANCE LIMIT, KSI
Bar chart presentation showing relative effects of various machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-4V.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys,"in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr .• Ed., American Society for Metals, Metals Park OH, 1982, P 354
439
440
17-27. Ti-6AI-4V: Effects of Various Metal Removal Operations on Endurance Limit Ti- 6Al-4V ANNEALED, 35 R c Surface Rouglme9s, AA
Gentle
SURFACE GRIND
I
14
67
Il
160
ECM FRONTAL
161
I 40
ECM TREPAN I
o
lO
J 40
I
60
I
I
80
ENDURANCE LIMIT, KSI
Bar chart presentation showing effects of various metal removal operations on the fatigue characteristics of titanium alloy Ti-6AI-4V.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in MachiningTitanium Alloys."in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 355
17-28. Ti-6AI-4V: Effect of Texture on Fatigue Strength EFFECTS OF TEXTURE ON SMOOTH FATIGUE LIFE
130 120
. ..
... 110 I
<,
E
b lOa
Load axis
.....
~
II [loio la
........ ........................
90
80 L--
Ti-6AI-4V Re-X Anneal R =0.1 .L..-
........ .....L
--------J
Nt - cycles
SON curves showing the effect of texture on the fatigue strength of Ti6AI-4V. Fatigue strength is greater when the stress axis coincides with the direction of a high density of basal poles.
Source:J. C. Williams and E. A. Starke.Jr., "The Role of Therrnomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys," in Deformation, Processing, and Structure, George Krauss, Ed., American Society for Metals, Metals Park OH, 1984, P 334
441
442
17-29. Ti-6AI-4V: Effect of Complex Texture on Cycles to Failure Ti-6AI-4V ~
900
E
-o~-
\
::;:
.,
-[]o-
\ \
z
800
-0
\
\
-, <,
~
.-
a. E
700
«
II> II>
'"
Air
...... ....
-- - - - - - - -
600
10 4
10 5 106 Cycles to Failure
10 7
S-N curves showing the effect of more complex texture on fatigue strength of Ti-6AI-4V. These data show that a mixed texture lowers the high-stress end oftheS-N curve preferentially.
Source: J. C. Williams and E. A. Starke, Jr., "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys, "in Deformation, Processing, and Structure, George Krauss, Ed., American Society for Metals, Metals Park OH, 1984, p 335
17-30. Ti-6AI-4V: Effect of Texture and Environment on Cycles to Failure TI-6AI-4V
-,
AIr
-, <, ............
_-----
lOS
Cycles to Failure
a
Q.
E
..
31/2 %
Ti-6AI-4V
900
IJ)
800
\
\
\
\ \
700
Nael Solution
\ \
-,
-,
-, <,
600
~------
104
105
106
10 7
Cycles to failure
b O! + ,8- processed Ti·6Al·4V, showing the effects oftexture and environment on fatigue strength. (a) Tested in air. (b) Tested in 3'12% NaC\, These data show that testing in an aqueous 3'12% NaCI solution reduces fatigue strength when the stress axis is along [0001].
SoN curves for
Source:J. C. Williams and E. A. Starke, Jr.. "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys," in Deformation. Processing, and Structure, George Krauss, Ed .. American Society for Metals, Metals Park OH, 1984, p 335
443
444
17-31. Ti-6AI-4V: Fatigue Crack Growth Rates
Kmin+lJ.KalK 1c
.:»:"
1 ..,z .....
..,e o '"
10- 10
L-_ _...L<...
_
log
lJ.K-
Schematic plot showing characteristic shape offatigue crack growth rate (doldN) versus cyclicstress-intensity (~) curves.
It can be seen that at higher growth rates there is a linear portion of the curve. 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. Most structural materials show variations in near-threshold FCP rate and in tiK,,, but fewer show significant variations in FCP rate in the Paris law regime. In contrast, Ti alloys show significant variations in FCP rate over the entire range. At the highest crack growth rates shown above, the FCP rate curve bends upward. This is controlled by fracture toughness. However, since crack growth rates are uncontrollably rapid in this latter regime, it is of little interest and will not be discussed further here. Moreover, since the majority of the lifetime of a crack component is spent in the low-FCP-rate regime, factors which control FCP at rates less than ~ 10-6 tu] cycle are probably most important. These factors include microstructure and texture.
Source: J. C. Williams and E. A. Starke, Jr., "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys," in Deformation, Processing, and Structure, George Krauss, Ed., American Society for Metals, Metals Park OB, 1984, P 338
17-32. Ti-6AI-4V: Fatigue Crack Growth Rates for ISR Tee, and Extrusions 10-
2
rr========::::::;,-----, Ti-6AI-4V R = O.l,ROOM TEMPERATURE TL AND LT ORIENTATION • ISR TEE o EXTRUSION
W ...J
000
o
U
>-
~ :r: o
z
z ~ "tl w· I-
« a: :r:
~
a
a: (9 ~
o « a: o w
::J (9
~
«
lL
o o
0
-7 ':------:~--........-_:';:-.......-7:........~::_":~
10 10
20
40
60
80 100
STRESS INTENSITY FACTOR RANGE.~K. KSI viN.
Fatigue crack growth results for ISR tee, and extrusions.
Data for both TL and LT orientations are shown above, along with data for the extrusion. In comparing individual results, no differences were seen between TL and LT. In the chart, 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. This is probably due to the extrusions being beta formed while the ISR tees are alpha-beta formed.
Source: W. T. (Ted) Highberger, Govind R. Chanani and Gregory V. Scarich, "Advanced Titanium Metallic Materials and Processes for Application to Naval Aircraft Structures," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, P 124
445
446
17-33. Ti-6AI-4V: Fatigue Crack Growth Rates
;
~
: : (INCHES/CYCLE)
0
I.IE 0
~ _°11
~ -
o
,~
o
'§ o
10
100
dK (KSI {INCH)
0- HEAT 3
ENVIRONHENT - LOW-HUHIOITY AIR
b - HEAT 2
ORIENTATION - RW
0- HEAT 1
R FACTOR
-to. 30
HEAT TREATHENT - RECRYSTALLI ZATI ON ANNEAL
Fatigue crack growth rates for three different heats of Ti-6AI-4V titanium alloy.
The fatigue crack growth rate in the RW orientation for this alloy, when recrystallization-annealed, behaved similarly with decreasing oxygen and aluminum. The crack growth rate is shown as a function of 6.K tested at an R factor of +0.30.
Source: M. J. Harrigan, M. P. Kaplan and A. W. Sommer, "Effect of Chemistry and Heat Treatment on the Fracture Properties of Ti-6AI-4V Alloy," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982,P 65
17-34. Ti-6AI-4V: Effect of Final Cooling on Fatigue Crack Growth Rates ~K, MPa •
-.jm 100
10 I
CONSTANT AMPLITUDE FATIGUE: R = 0.1; 10 TO 20 HZ MAXIMUM SCATTER OF ACTUAL DATA POINTS FROM ANY CURVE IS LESS THAN 40% WQ AC FC
WATER-QUENCHED AIR-COOLED FURNACE-COOLED
A'A
MILL-ANNEALED (DATA IN- ~ CLUDED FOR COMPARISON) FC WQ
k~ '!:/lAC
10.3
'"
o >.. u
':-
.s
100
I
Y
:/
10- 4
I
:'/
:/
Z
1I
~ "U
:/
:,:"
,.":'
10. 5
I 0.1
10
~K
I
ksi-
100
Jin":
Effects of final cooling rate on fatigue crack growth rate in duplex-annealed Ti-6AI-4V, I-in. plate, 1775 OF (968.3 0q, 1/2 h, air cooled; and 1450 OF (787.4 "C), I h, cooled as noted.
From the data presented above, it can be seen that air cooling, per se, produced little or no change in the cyclic crack growth compared to the mill-annealed base (22 material. The slightly decreased crack growth rates above a IJ.K of 20 ksi MPa· VITi) are, more probably than not, the result of the higher fracture toughness of the air-cooled material. However, both water quenching and furnace cooling resulted in fatigue crack growth rates noticeably different from those measured for the mill-annealed base material. As shown, furnace cooling had a consistently detrimental effect on the crack growth rate while water quenching produced $!eatly increased crack growth rates above a stress-intensity range of 18 ksi yin. (20 MPa . ~). The accelerated growth rate above 18ksi \!Ill. (20 MPa . may be attributed to the proximity of the maximum stress intensity to the critical value. The critical stress-intensity value for water quenching was an exceptionally low 38 ksi .jll1. (42 MPa . Jill).
Jill.
Jill)
Source: R. E. Lewis, J. G. Bjeletich, T. M. Morton and F. A. Crossley, "Effect of Cooling Rate on Fracture Behavior of MillAnnealed Ti-6AI-4V," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 90
447
448
17-35. Ti-6AI-4V: Effect of Dwell Time on Fatigue Crack Growth Rates 10- 4 ,.--
~
u >u
-
-
-
-
10- 5
.... Q>
(
.s
I
a.
~ .J::.
0,
-
-
-
-
...,
I
I I I I
Q>
'i0
-
10- 6
-" u
m
U
-
Sinusoidal loading
--Dwell at maximum load
10-7
100
10 Stress intensity factor range (MPa m 1/ 2 )
A phenomenon which may be unique to certain titanium alloys is the effect of dwell periods at maximum load on rates of growth of fatigue cracks. This effect is shown schematically here, and increases in the rate of crack growth of as much as 50 times may occur compared with results obtained in tests on the same material subjected only to sinusoidal stress cycles. Dwell effects are maximized in alloys containing substantial amounts of the a-phase which have a preferred texture such that stressing is normal to the basal planes, whereas they appear to be insignificant if stressing occurs parallel to the basal planes of the aphase, or if the microstructure is homogeneous and fine grained. Particular attention has been paid to a / {3 alloys, e.g., Ti-6AI4V, in which dwell effects have also been found to decrease with increasing amounts of the {3-phase in the microstructure. In all cases, dwell effects disappear when stressing occurs at temperatures above 75°C (165 "Fj.and they are generally considered to arise from the preferential diffusion of hydrogen, during the dwell period, to regions of localized hydrostatic tension ahead of an advancing crack. Such an accumulation of hydrogen would tend to embrittle this region, and it has even been suggested that brittle plates of TiH 2 may be formed.
Source: I. J. Polmear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH, 1981, P 200
17-36. Ti-6AI-4V: Fatigue Crack Growth Data 0
• Annealed 2 hours at 705 C. air-cooled after forging transus 1005 C " Axial loading: smooth specimens. K = 1.0 0
• alP
t
r
~
I -. ", .'
" ..
Ti-6AI-4V
-SA
10
20
50
100
Left: Fatigue crack growth rates for Ti-6AI-4V rolled plates in the ,a-annealed (HA) and mill annealed (MA) conditions. BA = 0.5 h 1038 0 C, air-cool to room temperature. Tests conducted at 5 Hz using compact tension specimens. Ratio of minimum to maximum load = 0.1. Above: Branching of fatigue cracks within the Widmanstiitten packets of the a-laths.
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). This effect, in turn, is attributed to the slower progress of cracks through the Widmanstatten microstructure, particularly at stress intensities below a critical value at which desirable crack branching occurs within packets ofthe a-laths.
Source: I. J. Polmear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH, 1981, P 179
449
450
17-37. Ti-6AI-4V P/M: Comparison of HIP'd Material With Alpha-Beta Forgings for Cycles to Failure r:
70
~o o
~OHIP
o
------00------
40
Upper and lower limits of
~-_-:--.. alpha-beta-processed forgings
as listed in AFML TR-73-301 30 LJ...Ll..lll_--.J_L-L.l.....cr::ti±==I==:±=~:::l.-'~ 106 10 Cycles to failure Notched fatigue strength of HIP'd P 1M Ti-6AI-4V compared with fatigue strength ofalpha-beta processed forgings.K, = 3; Hz = 30;R=
0.1.
Source: J. H. Moll, V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications," in Powder MetallurgyApplications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 286
17-38. Ti-6AI-4V P/M: Comparisons of HIP'd Material With Annealed Plate for Cycles to Failure 140
r
120~ 100
'GM"S'$, K'S/
o
80 ~ Q I/IPC'tCl.£ o 15S0i:-I-5 1t:.'6/-II-IR
40t
• a
''''50.~-1'51t:'8/- 3/-1J:?
22'50·~-I0/t:!9I-II-IR
ANN P!AT~
I
20 1----,-_ _--'--_ _---"I 10~
/0'-
10-5
-'-:-_ _----', 1
/0·
/0'
C~ESTO~(L~ SoN curve for HIP'd Ti-6AI-4V and annealed plate. According to the above data, fatigue results for Ti-6AI-4V are within the required range for plate properties from MIL-T-9046.
Source: W. Theodore Highberger, "Manufacture of Titanium Components by Hot Isostatic Pressing," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, P 304
451
452
17-39. Ti-6AI-4V P/M: Effect of Powder Mesh Size on Fatigue Properties 70.3 (100) 56.2 (80) 'Vi ~ m
a..
42.2 (60)
:E
VI' VI Q.l
....... en
28.1 (40)
o As received powder (SM 772) 0-80 mesh screened powder (SM 768) 14.1 (20) 0
104
·10,
70.3 (100 )
56.2 (80 ) 'u;
-:-- r----
~
~
42.2 (60 )
li.
--
~li. Pli.
Ali.
o~l
:E
....~'
28.1 (40 ) O As received powder (SM 772) li. -80 mesh screened powder (SM 768)
en
14.1 (20 )
io,
10,
Top: Room-temperature properties. Bottom: Properties at 700 OF (370°C).
High-cycle fatigue (HCF) data were developed on Ti-6AI-4V (Std) by Williams International in a program to apply near-net-shape HIP technology to a compressor rotor part for the F-I07 cruise missile engine. In this study, two size fractions of powder were used: -35 mesh (as-received) and -80 mesh. There was no difference in HCF test results between the two sizes. Roomtemperature and 700 OF (370°C) S/ N curves are shown above.
Source: J. H. Moll. V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications,"in Powder MetallurgyApplications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 286
17-40. Ti-6AI-4V P/M: Comparison of Blended Elemental, Prealloyed and Wrought Material for Effect on Cycles to Failure 1200
1000
140
co
Q..
'iii
~
iii III
.><
iii
800
III
e li)
~
100 ti
E :J E
E :J E 600
'x co
'x co
~
~
60
400
200 10'
10'
10' 10' Cycles to failure
10'
10'
SON curves showing comparison of smooth axial fatigue behavior of Ti-6AI-4V blended elemental and prealloyed P 1M compacts with wrought annealed material. Tested at room temperature, R = O.
The fatigue behavior of titanium PIM compacts is compared to wrought products in the graph above. The blended elemental material is inferior to prealloyed compacts and II M materials. This is caused by residual chlorides and consequent porosity; also, chemcial heterogeneity may lead to areas of similarly aligned alpha plates. Blended elemental compacts, however, compete well with many titanium alloy castings in fatigue strength. Prealloyed powder compacts exhibit fatigue behavior equivalent to that of Ij M materials. This situation is achieved by careful control of cleanliness (powder handling) and microstructure. Cleanliness depends on the environment in which the powder is produced, conditions of subsequent handling, and microstruture developed by compaction. Cleanliness dictates the amount of contamination contained in the final product; microstructure determines the ability of the compact to accommodate foreign particles and resist crack initiation.
Source: Metals Handbook. 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,P 753
453
454
17-41. Ti-6AI-4V: P/M Compacts vs 11M Specimens: Cycles to Failure co
Q..
~
1200
'"
1000
en
~
1ii
800
E ::J E
600
x
CO
~
400 103
'(ij .>t:
10
5
106
7
10
10
f/l
~f/l
100
E ::J E
60 104
'"
140
8
'x CO ~
Number of cycles to failure Fatigue chart presentation showing a comparison of fatigue behavior of Ti-6AI-4V compacts with ingot metallurgy material.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, P 44
17-42. Ti-6AI-4V: Comparison of Specimens Processed by Various Fabrication Processes for Cycles to Failure 1200 Axial fatigue smooth Room temperature R = 0,1 Annealed
ro
a.. ~
160
120
800
'ijj ..".
vi
u>
III
III
~ 1;;
~ 1;;
E ::J E
80
E ::J E
'x ro
'x ro 400
~
~
40
o"10
-'-
...L...-
--'
3
....L..
---I0
108 Cycles to failure
SON curves (bands) for titanium alloy Ti-6AI-4V processed by various fabrication processes. The inconsistentfatigue life ofthe hot isostatically pressed product is usually casued by inclusions in the compact.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,P 439
455
456
17-43. Ti-6AI-4V: Comparison of Fatigue Crack Growth Rate, P/M vs 11M Stress intensity, ksiVTrl.
10
3
10-4
1
10 11M and P/M Ti-6AI-4V at room temperature (laboratory air), R = 0.1, at 5 to 30 Hz
102
10-2
Q)
u
Q)
~ c
>
10- ~ E E
......ai
i...
3
U
ai
III
.s::. 10- 5
~
Recrystallization anneal
s:
~
...0
...001
10- 4 ~
.>0:
(J
(J
...
... o
III
III
U
10- 6 10- 5
Stress intensity, MPaViTl Comparison of fatigue crack growth rate of Ti·6AI·4V P 1M compact with 11M material heat treated to various conditions. The fatigue crack growth rate of blended elemental and pre alloyed compacts is equivalent to 11M material with the same microstructure.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, p 752
17-44. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure
130
~
<,
<,
110
0---,
0
<,
l!l'
c: <,
u;
>l
TEST CONDITIONS: CONSTANT AMPLITUDE, R ~.1 K ~ 1.0 T STRESS RELIEVED, FLUSH WELD BEAD
90
w
a:
0
~
t-
IJ)
:E :::> :E
x
«
"""-
CD
~- ~
70
:E LEGEND
50
---/:::,.
BASE METAL
--0
SSEB WELDEO
10
6
NUMBER OF CYCLES (LOG SCALE) TO FAILURE
S-N curve for titanium alloy plate-base metal versus SSEB-welded. Results show that the constantamplitude fatigue life of SSEB weldments in O.440-in.-thick plate equals that of the base metal.
Source: R. H. Will, J. G. Madora and H. P. Ellison, "Sliding-Seal Electron-Beam Welding of Titanium," in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American SccietyIor Metals, Metals Park OH, 1981, P 87
457
458
17-45. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure
120
TEST CONDITIONS: CONSTANT AMPLITUDE, R KT~
~.1
10
STRESS RELIEVED, FLUSH WELD BEAD
1lX'
o
80
o ~
LEGEND:
40
--- 0
SSEB WELDED
---- 6
BASE METAL
20
SON curves for titanium alloy plate-base metal versus SSEB-welded. Results show that the constantamplitude fatigue life of SSEB weldments in O.940-in.-thick plate equals that of the base metal.
Source: R. H. Witt, J. G. Madora and H. P. Ellison, "Sliding-Seal Electron-Beam Welding of Titanium," in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, p 87
459
17-46. Ti-6AI-4V EB Weldments: Base Metal Compared With Flawless Weldments 140
0
0
o
O.OSQ-IN.·THICK BASE·METAL Ti-6AI-4V STOA CURVES
0
120
ci
s"
o
100
o
0
en
o
~
en' en w a: Ien
«X
::E
o
SO
o
60
~
40
20 103
10 5 CYCLES
SoN curves for EB weldments that were flawless (lower two curves). Upper curve shows scatter band for base metal (O.080·in.-thick Ti-6AI-4V STOA.).
Source: R. Witt, A. Flescher and O. Paul, "Weldability and Quality of Titanium Alloy Weldments," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982,P 313
460
17-47. Ti-6AI-4V EB Weldments: Effects of Porosity on Cycles to Failure 140
a a 120
a
O.oaO·IN.·THICK aASE·METAL Ti·6AI·4V STOA CURVES
a
a
~
d
100
II
a:
in
~
~'
ao
w
a:
Ii; X
:E
60
40
20 103
105 CYCLES
SON curves for ED-welded Ti-6AI-4V titanium alloy showing effects of porosity.
Above are shown experimental data obtained for porosity-containing ED welds which are superimposed on a set of curves for the base material (0.080-in.-thick Ti-6AI-4V STOA sheet) at various K, factors. For the points within the boundaries of the band, radiography indicated scattered porosity (0.003 to 0.005 in. in diameter). For points below the lower boundary of the band, radiography indicated either linear or heavily scattered porosity.
Source: R. Witt, A. Flescher and O. Paul. "Weldability and Quality of Titanium Alloy Weldments," in Titanium and Titanium Alloys: Source Book Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 312
461
17-48. Ti-6AI-4V Gas Metal-Arc Weldments: Effects of Porosity on Cycles to Failure 140
D
0.250-IN.-THICK BASE-METAL Ti-6AI-4V STOA CURVES
D
120
0 ~ 100
D
Cii ~
iii lI) w
a:
BO
lll)
~
::E
60 D
40
D
o
D
A
o 105 CYCLES
SON curves for porosity-containing gas metal-arc welds. In the above graph the experimental fatigue data for porosity-containing GMA W weldments are superimposed on SoN graphs for Ti-6AI-4V STOA material (0.25 in. thick) for various K 1 factors.
Source: R. Witt, A. Flescher and O. Paul. "Weld ability and Quality of Titanium Alloy Weldments," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, p 313
462
17-49. Ti-6AI-4V: Unwelded vs Electron Beam Welded Material for Cycles to Failure
III 0..
en III w
It:
t;
---
100 80
60
.
-'"
..
'r.I
~
.. Q=
(!)
z
5z
40
S
20
o
It: W C[
f· 7500 CPM Kr
HEAT TREATED TI-6AI-4V
• HEAT TREATED ELECTRON BEAM WELDED a STRESS RELIEVED TI-6AI-4V
=1.0
I 6 10 CYCLES TO FAILURE
Room temperature rotating-beam fatigue life of unwelded and electron-beam-welded Ti-6AI-4V titanium alloy in fully heat-treated condition. Decrease in fatigue strength ofthe weldment relative to the parent metal did not exceed 12%.
Source: S. M. Silverstein, V. Strautman and W. R. Freeman, "Application of Electron Beam Welding to Rotating Gas Turbine Components," in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, P 169
17-50, Ti-6AI-4V: S-N Diagram for Laser-Welded Sheet 120
..... .... ci
Legend
$1!!~iJfift:
100
II
a: ..... 'iii
.¥
.....
'@
80
Ul Ul
E E :;,
oco
.0
-,
Gl Ul
60
<,
')(
IV
::iE
o 0.230 in, (0.584 cm) sheet • 0.140 in. (0.356 cm) sheet Mean curve for plasma-arc welds 700 co..... I 0 of mean fatigue strength .... of all un-welded control )( 600 N specimens
40
20
~
0
<,
<,
-..... E
•
---
c
----
Lower bound data for pIa welds J on flat sheet with filler 10 6
4 runouts
.....
l3Gl Ul
E :;, E
400
')(
----
10 7 Cycles to failure
IV
300
::iE
10 8
S-N diagram for laser-welded titanium alloy sheet,
The fatigue properties of welds as shown above indicate that under proper welding conditions, laser welds can be made in Ti-6AI-4V which exhibit base metal fatigue characteristics. The best laser weld failures initiated at sites in the base metal, whereas other weld failures originated at undetected small pores. Where failures initiated in the base metal, it was concluded that no porosity or weld defects of sufficient size to preferentially initiate fatigue fractures was present.
Source: E. M. Breinan, C. M. Banas and M. A. Greenfield. "Laser Welding-The Present State-of-the-Art," in Source Book on Electron Beam and Laser Welding. Melvin M. Schwartz. Ed., American Society for Metals, Metals Park OH, 1981, P 289
463
464
17-51. Ti-6AI-4V (Cast): S-N Diagram for Notched Specimens 1000 900
o
0
800 t- o
..
700
:E
600
Do ~.
...'" ~
E ::J E
.
'j(
:E
500 400 300 200
0
120
oooeJO
- 90
0000y 00 0 0 00'17001:>.61:>.
- 60
~f:>6'17 00
+000#
o 00+-tp.~66 '17 'Ii' '17 o 000 01:>.Q) +0 x ooo,+°oOOx O +0 0 o-tCo 00
00
'';;
.>t
.n
...e'"'" E ::J E
..
'j(
:E
+
0 0
0
+o~ooo
30
oc:P
100
No. of stress cycles
Notch fatigue strength of as-cast Ti-6AI-4V. Each symbol represents fatigue data from a different source. Stress ratio,R, typically was +0.1; stress concentration factor, K/, was mostly 3.0, but a few tests were run at K/ = 1.0.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose Metals, American Society for Metals, Metals Park OH, 1980,P 411
18-1. Zirconium 702: Effects of Notches and Testing Temperature on Cycles to Failure
'"0..... N
e U
-,
Cl
-"
to
en en
w a::
l-
en
0.5
CYCLES S-N curves for zirconium grade 702, showing effects of notches and elevated temperature (400°C, or 750 OF) on fatigue characteristics.
As indicated above, zirconium and its alloys exhibit a fatigue limit behavior similar to most ferrous alloys.
Source: Donald R. Knittel, "Zirconium." in Corrosion and Corrosion Protection Handbook, Philip A. Schweitzer, Ed., Marcel Dekker, Inc.• New York NY, 1983. p 198
465
466
19-1. Steel Castings (General): Effect of Design and Welding Practice on Fatigue Characteristics
'W:....
S-N CURVE -CYCLES OF
X "" STRESS VS COMPUTED
~
"" .... x STRESS .... .... ~ x Xx 00 .....x ~~~
::f ....0.1:-:!-::
o ...............
x~x
x
10 6
107
NUMBER OF CYCLES
S-N curve for cast box designs. u,
o
~
50
(5.Q.~ ~ I 40 Vl
llJ
[3 ~ 30 11::11-0 VI Cl:
olE llJ
~
-I ::::I
3
(3
x,oX,
S-N CURVE -CYCLES OFSTRESS COMPUTED STRESS
'o'x VS '-.""', '"'ox
lbX~', "
X
"'80x.. .
XX x -_ --o-J!l
20
0",,--
-0_
0_
10
0
X - 2-BW HAND WELDED, STRESS RELIEVED-o - 2 - BW MACH WELDED. STRESS RELIEVED-
105
106 NUMBER OF CYCLES
SoN curves of box weldments, comparing hand weldments with machine weldments. All weldments were stress relieved at 1100 F (593 Qq. Q
The S-N curves shown above indicate that: (1) the welding practice is of no great importance; and (2) the cast steel box design is superior to a weldment design.
Source: Steel Castings Handbook. 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 7-6
19-2. Steel Castings (General): Effects of Discontinuities on Fatigue Characteristics 1.0
0.9
0.6
0.7
0.6
0.5 :I: l-
:I: I-
W
W
lI)
lI)
W
W .J
e> ~ 0.4 Z
It: It: l- I-
~
e>
~
~
AVERAGE YIELD STRESS
RANGE
TENSILE STRENGTH
0.3
0.2
0.1
0 0.5
MEAN - 0.1
0.6
STRESS
TENSILE STRENGTH
- O. 2
- 0.3
-0.4
Goodman diagram for bending fatigue for normalized and tempered 8630 cast steel. (Machined notch of R. R. Moore specimen: 60° included angle, 0.0015-in. (0.0381mm) root radius.]
Surface condition has a significant effect on fatigue life and fatigue limit. A highly polished smooth test specimen can exhibit twice the fatigue strength of a rough machined sample. A good design approach is to use the notched fatigue limit asa design value. For cast steels a O.OOIS-in. (O.0381-mm)root radius circumferential notch in a rotating beam fatigue specimen reduces the fatigue limit by about 0.7 of the unnotched value. This is sufficient to account for variations in surface finish and minor surface discontinuities. The above diagram shows that even severe surface discontinuities, not normally permitted by workmanship standards, do not reduce the fatigue limit by much more than the 0.7 value. The above emphasis on surface discontinuities is due to the fact that subsurface discontinuities which do not have a crack-like sharpness and which do not significantly reduce the load-bearing area of a component generally have little effect on fatigue performance.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980, P 7-6
467
468
20-1. Closed-Die Steel Forgings: Effect of Surface Condition on Fatigue Limit 9oor--....,._~---------,r------------,----------~
125
Ground and polished 8oot-----.::>.,;~~
.......:::_II__-___7'''-------_+---------~
7oo~------~......,;~~-..-==-"""'-"""",------,,L----+----------d
100
6oo~--------___I--=~:::""....,--=_=-=""'l-"",""----------l
~
500 ~--------___I---___;.,L_-=....,.-=.......; ::_t--"'-..=_-:=.""""=__--=l 75 ~
~ 4oot-----------1------------'''''f.,=---'''......, -----'''''-..---=c.j
Ul
Ii
l' 50 Ul
3oot-----------1----------+----"'-,--=---~c.j 2oot-----------1----------+---------''''''''',----~c.j
25
loot-----------1----------+------------''-l 0'-------------'------------'-----·-------' 0 103 106 Number of cycles to failure
Tensile strength, ksi
50
100
150
1000
I II
v
800
MaChine[5h
.
D..
::;;
600
-:V
] '!l
.~ 400
o
200
125 ~
100
75
/ --- >-_L --1,....--t.> ......
"....
400
600
800
1000
1200
50
----1---
1400
~
] ~
.~
As-forgedor
decarburized
u.
200
250
200
1600
-... 1800
.;;;
u.
25
2000
Tensile strength, MPa
Application of small-scale laboratory fatigue testing to the analysis of components or assemblies introduces additional variables. One is the effect of surface condition. The curves in the top curve above demonstrate that the fatigue strength of steel specimens varies markedly, depending on whether the surface is polished, machined, hot rolled, or as-forged. The steel tested was an unidentified wrought low-alloy steel heat treated to 269 to 285 HB, equivalent to a tensile strength of 876 MPa (127 ksi) and a yield strength of 696 MPa (101 ksi). Sample preparation required that the specimens be machined and polished after heat treatment and that rolling or forging precede heat treatment. For a fatigue life of one million cycles, the fatigue limit was 393 MPa (57 ksi) for the ground specimens, 317 MPa (46 ksi) for the machined specimens, 207 MPa (30 ksi) for the as-rolled specimens, and only 152 MPa (22 ksi) for the as-forged specimens. The curves in the bottom graph apply to steels with tensile strength ranging from 345 to 2070 MPa (50 to 300 ksi) and are approximations from several independent investigations. Sample preparation for "as-forged or decarburized" specimens at the 965 MPa (140 ksi) tensile-strength level include 4140-type steels rough machined from bar stock, heated to approximately 900°C (1650 OF) in a gas-fired muffle for 20 to 30 min, very lightly swaged from an original 7,47-mm (0.294-in.) diameter to a final diameter of 7.16 mm (0.282 in.), and air cooled. Heat treatment consisted of austenitizing in a salt bath at approximately 830°C (1525 OF)for 45 min, oil quenching, tempering in air for 1 h at approximately 620°C (1150 OF), and water quenching. Forging and heat treating produced a surface decarburized to a depth of about 0.06 mm (0.0025 in.). These specimens exhibited a fatigue strength, at 106 cycles, of about 310 MPa (45 ksi), compared with 470 MPa (68 ksi) for samples that were not forged but were machined or polished and free of decarburization. Decarburization lowers the strength levels obtained by heat treatment. Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 355
21-1. P/M: Relation of Density to Fatigue Limit and Fatigue Ratio
'" 50 ~ (345) tJl 40 0.. (276) o g 30 _ (207)
'E :.:::i
20
~ (138) Cl
~
10
(69)
.50 0
~
tx: Q) ~
.45 .40
Cl
~
u..
.35
6.4
6.6
6.8 7.0 Density, gr/cucm
7.2
The relationship of fatigue strength to density is shown above. Fatigue strength is best at high densities. For similar P / M and wrought parts, the ultimate tensile strength to fatigue strength ratios are the same. However, fatigue strengths of P/M parts generally are more stable and uniform than for wrought parts. Parts containing nickel show improved fatigue resistance compared to iron-carbon steels, and high-density nickel steel parts can be case hardened to improve wear and fatigue properties.
Source: Kurt H. Miska, "Powder Metal Parts," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 3
469
470
21-2. P/M: Relation of Fatigue Limit to Tensile Strength for Sintered Steels N/mm
2
150
•
0
•
00
•o
100 o o
200
400 Tensile strength)
500
2
600
N / mm
Fatigue limit of different sintered steels as a function of tensile strength. Triangles are values for materials without phosphorus; open circles correspond to PNC materials, closed circles to PASC materials.
Source: Per Lindskog, "The Effect of Phosphorus Additions on the Tensile, Fatigue, and Impact Strength ofSintered Steels Based on the Sponge Iron Powder and High-Purity Atomized Iron Powder," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH. 1979, P 46
21-3. P/M (Nickel Steels): As-Sintered vs Quenched and Tempered for Cycles to Failure
/
40
r
0,1 Quenched and Tempered _ Tensile Strength 105,000 psi
As - Sintered, Tensile Strength 67,OOOpsi
10
106
7
10
Cycles to Fuilure , N. SON diagrams representing fatigue behavior of7.0 g/ cm 3 density, 4 Ni-0.48 C steels, and effect of quenching and tempering on tensile and fatigue strength.
One of the characteristics of the fatigue behavior of wrought steels is that the S- N curve usually shows a distinct fatigue limit. This is most marked in wrought plain carbon steels and usually occurs between 105 and 107 cycles. A typical SoN curve for an as-sintered nickel steel is shown above. As-sintered nickel steels possess distinct fatigue limits occurring between 106 and 108 cycles.
Source: A. F. Kravic and D. L. Pasquine, "Fatigue Properties of Sintered Nickel Steels," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 28
471
472
21-4. P/M (Nickel Steels): Relation Between Fatigue Limit and Tensile Strength for Sintered Steels 60
r---,...-----,---,----...------,---,---....,.--=__--, LEGEND
50
•
Smooth As-Sinfered
0
Smooth Quenched
a
Tempered
.... 2.2 Kt Notch As-Sinlered t:J. 2.2 Kt Notch Ouenched-B Tempered
Ui 40 (L
o o o I ~
E 30
--'
., :::J
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.
A plot ofthe fatigue ratio (above) indicates an average smooth value of 0.4 up to 150,000 psi tensile strength. Thus the averagefatigue ratio for sintered nickel steel is 0.4 which is apparently independent of density level, alloy content, and state of heat treatment and therefore can be used to predict the fatigue behavior of other sintered nickel steels.
Source: A. F. Kravic and D. L. Pasquine, "Fatigue Properties ofSintered Nickel Steels," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 30
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---'''''
40
Smooth
2.2 Kt Notched
10
106
107
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
473
474
21-6. P1M (Nickel Steels): Effect of Notches on Cycles to Failure for the Quenched and Tempered Condition
0-0-'in
0--
a.
o 30
o o
2.2 Kt Notched
If)
~
e
20
+ If)
10
106
107
Cycles to Failure, N.
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
40
'" a..
'" '" e
l""- I'--.
Nilrided
Nitriderl
30
= = =-
en
--.........
r-
to-.
Not nitrided r--
20
-
r---..r--
10 lQ5
Smooth (K=1)
Notched (K=2)
I I III
I IIII
Not nitrided
10 8 lQ5
Cycles to Failure S-N curves for sintered powders (low-carbon; 1 to 5% copper, 7.1 g/cm 3 density). As shown above, notches greatly lower fatigue strength, particularly of those that were not nitrided.
Source: "Nitriding Improves Fatigue Resistance of P 1M Parts," in Source Book on Nitriding, American Society for Metals, Metals Park OH, 1977, P 292
475
476
21-8. P1M (Sintered Iron, Low-Carbon, No Copper): Effect of Density and Nitriding on Cycles to Failure
<; I I III I I II I III r
30
Density - 6.4 g per eu em "iii D.-
I II Density - 7.0 g per eu em
Nitrided
20
r-..... r--...
o
0 0_
""" t--.....
<11
~
en 10
-
Not nilrided
Nilrided Not nitrided
11 10 6
10 8 !OS
10 6
eye les to Fa ilure 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
I"":iZ ~
~'''':': :::
'Vi
a.. 40
= =. ~ =
~
Sintered Iron
::fit k:t
I
Nitrided .:.:.:.:.:.: ;:::;:::::; :::;::;:: ;:;:; ::: :: :.:.:.:.:.:.
~
::::; :::::'~ 00
::::::;:;::: :.:.:.:.:.: ~:::::: ;~~~;~~~~~~ U~
Not nihided
~ ~ ~ mmmm ~~t~r ~~r m= ~
~'" .....:....
_:
20
10 6
10 7
00
-
10 8 10 5
.....':':":
~ ::::;:
Nilrided :;:;:;:;::::: ::: :~ .........• :.:~:~:~:~:~: ~:~:~ :~:~ ~:
m it ~ : mIt ~~tt :~t~ ~~~~ """" <' t
106
Not nilrided
::: ~~ ~~~~~~~~~~~~~ ~~~~~~~~~ ~~~~;~ .':' ::::::::: t~~ jtm~~ :::::;:;: ~ ~ ~: ~m ~ 10 7
108
Cycles to Failure Left: Effect of salt bath nitriding on fatigue strength of ductile iron. Right: Effect ofsalt bath nitriding on P 1M parts. Specimens were made from a 3% copper sintered iron ranging from 6.2 to 7.0 g/cm l in density. All specimens were unnotched and were heated in a nitriding salt at 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
477
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
• staintess clad mild steel steet 0 Mild sleel
0____~ .
' Slainl e s s
•
2 /5 -e
12-5
'"
~
0 o~~-'====t
=
10 '1:
.e
_
0_
OJ'
"l::I
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
479
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, c
~
1-------
300
Cl>
.~ 200
'"
u,
100
o
SAE 1000
SAE 1015
SAE 1035
SAE 1045
SAE 1060
600
'"
500
0-
:2
-S' 400
'"
c
~
-
300
Cl>
::l .g» 200
'"
u,
100 0
SAE 1035
SAE 5134
Nitralloy
Bar charts showing increases in fatigue limit that may be obtained by nitrocarburizing (gas or liquid processes).
The amount of improvement in fatigue strength of nitrocarburized materials, as determined with unnotched Wohler test specimens, depends on the hardness and depth of the diffusion zone. The potential for improvement in fatigue strength lessens with increasing carbon and alloy content.
Source: Metals Handbook, 9th Edition, Volume 4, Heat Treating, American Society for Metals, Metals Park OH, 1981, p 269
23-2. Carbon and Alloy Steels (Seven Grades): Effects of Tufftriding on Fatigue Characteristics
481
Normalized and tufflrided,
Tufftrided, 90 min, 1050 F, water-quenched
90 min, 1050 F, water-quenched "" 1---, ~ I------'r---.....:..;'--T''''''-=..:~..:...;.;.'''''''':....:l.::..:..;,;.:..;;;.;.'__, ~ ~
80
"is~
en
60 J----,---------t
.".
t-----t
+ I---l~-~----t
'"
c,
g 401-----1
=-
20 1-----11_
5134 SAE Basic fatigue characteristics are directly related to carbon content, as indicated in the above bar charts for carbon and alloy steels (unnotched test bars). Tufftriding these steels shows results which prove that fatigue strength increases inversely with carbon content; that is, the lower the carbon, the greater percentage increase in fatigue strength by Tufftriding.
Source: Edward Taylor, "Tufftride: Only Skin Deep"," in Source Book on Nitriding, American Society for Metals, Metals Park OH, 1977, P 280
482
23-3. Carbon and Alloy Steels (Six Grades): Effects of Nitriding on Fatigue Strength
60
After Atmoaphere NitridJng
After Hardening
111111/ I I
55
.... III
.c
..10:
I
11111
I
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
o
0.4
340
320
\
~
:2 £ 300
...en c
~ '" Q)
~
280
''::;
\
'"
0.8
1,2
1.6
45 'iii
.:.l
-,
J::."
-,
'"
u,
260
1» c
e
1;; Q)
<,
::s
en
40
r-, .....
'fJ
u.
~
240
o
10
20
3D
40
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
483
484
24-1. Coil Springs, Music Wire (Six Sizes): Data Presented Means of a Goodman Diagram
by
40 200 1-----,,(---+----+----+----+--_= 20
o Minimum stress,MPa
Wire diam
Spring
on
Spring No.
mm
in.
mm
1 2 3 4 5 6
0.81 0.81 1.22 2.59 3.07 4.50
0.032 0.032 0.048 0.102 0.121 0.177
9.52 6.35 15.88 22.22 22.22 22.22
Free length
in.
Spring index
mm
in.
Total turns
Active tums
Total tested
0.375 0.250 0.625 0.875 0.875 0.875
10.7 6.8 12.0 7.6 6.2 4.9
22.10 26.97 44.45 60.20 57.15 57.15
0.87 1.062 1.75 2.37 2.25 2.25
6.0 7.0 7.0 7.0 7.5 7.5
4.2 5.2 5.2 5.2 5.7 5.7
16 28 38 43 35 25
Data are average fatigue Iimita from S-Ncurv.. for 185 unpeened springs of various wire diameters run to 10 million cycles oCetrese.Allstresees were corrected for curvature using the Wahl correction factor. The springs were automatically coiled, with one tum squared on each end, then baked at 260'C (500 'F) for 1 h, after which the ends were ground perpendicular to the spring axis. The test load was applied statically to each spring and a check made for set three times before fatigue testing. The springs were all tested in groups of six on the same fatigue testing machine at ten cycles per second. After testing. the unbroken springs were again checked for set and recorded. Number 4 springs, tested at 1070 MPa (166 ksi) max strese, had undergone about 2,",%set after 10 million strese cycl.., but the stresees were not recalculated to take this into account. None of the other springs showed appreciable set. The tensile strengths of the wires were according to ASTM A228.•
By means of the Goodman diagram many fatigue-limit test results can be shown on the same diagram as indicated above. In this diagram, line OM represents the minimum stress for the cycle; the plotted points represent fatigue limits for the respective minimum stresses used. The vertical distances between these points and the minimum stress reference line represent the stress ranges. Some scatter may be expected, at least partly attributed to normal changes of tensile strength with wire diameter. Line UT is usually drawn to intersect line OM at the average ultimate shear strength of the various sizes of wire.
Source: Metals Handbook, 9th Edition. Volume I, Properties and Selection: Irons and Steels. American Society for Metals. Metals Park OH. 1978, P 293
24-2. Coil Springs: S-N Data for Oil-Tempered and Music Wire Grades 1 5 0 0 , - - - - - - - - - - - - - - , - - - - - - --r-r-- - - - - - - - - , 200 1250t---------+--------+------------j
175
~ ~
*
~
::i 1000f-------""....,,-=-::l-::--------+f 00
---l150 125
~
750t---------+----.::::!!oo""""'::-l----------l
500L ;--104
- - - - - '-;-- - - - - - --'-::-- - - - - - ___==! 105
106
Number of cycles to failure
la)
Type of wire
Number of springs
Oil tempered
6
Music wire
8
r
Avg
~
~ 104
105
Number of cycles to failure (b)
(a) Springs were made of minimum quality music wire 0.59 mm (0.022 in.) in diameter. Spring diameter was 5.21 mm (0.205 in.);D/dwas 8.32. Minimum stress was zero. Stresses corrected by Wahl factor. (b) Life of springs used in a hydraulic transmission. They were made of oil-tempered wire (ASTM A229) and music wire (A228). Wire diameter was 4.75 mm (0.187 in.), outside diameter of spring was 44.45 mm (1.750 ln.), with 15 active coils in each spring. The springs were fatigue tested in a fixture at a stress of 605 MPa (88 ksi), corrected by the Wahl factor.
The upper graph is a typical S- N diagram showing results of compression testing coil springs, where the minimum stress is zero while maximum stress is shown by points on the chart (see spring and testing details given in caption). The lower graph shows an alternate method of presenting fatigue data for steel springs.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, p 292
485
486
24-3. Coil Springs: Effects of Shot Peening on Cycles to Failure 100
90
80
t
<:
e "
70
'in <:
-
~
s
°
111
60
50 ~
40
~~ r-... <,~tressed
~
-----
/peened
in bending
""-Not peened I-
-
/Peened
~Stressed
in torsion
'\:"Not peened
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. Shot peening can be applied to wire 1.6 mm (1/16 in.) or more in diameter, and slightly smaller wire using special techniques. The kind of shot used is important; better results are obtained with carefully graded shot having only a few broken, angular particles. Shot size may be optimum at roughly 20% of the wire diameter. However, for larger wire, it has been found that excessive roughening during peening with coarse shot lessens the benefits of peening, apparently by causing minute fissures. Also, peening too deeply leaves little material in residual tension in the core; this negates the beneficial effect of peening, which requires internal tensile stress to balance the surface compression. Shot peening is effective in largely overcoming the stress-raising effects of shallow pits and seams. Proper peening intensity is an important factor, but more important is the need for both the inside and outside surfaces of the spring to be thoroughly covered. An Almen test strip necessarily receives the same exposure as the outside of the spring, but to reach the inside, the shot must pass between the coils and is thereby much restricted. Thus, for springs with closely spaced coils, a coverage of 400% on the outside may be required to achieve 90% coverage on the inside. Cold wound steel springs normally are stress relieved after peening to restore the yield point. A temperature of 230°C (450 OF) is common because higher temperatures degrade or eliminate the improvement in fatigue strength. The extent of improvement in fatigue strength to be gained by shot peening, according to one prominent manufacturer of cold wound springs, is shown in the above diagram.
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 297
24-4. Coil Springs, 8650 and 8660 Steels: Relation of Design Stresses and Probability of Failure 800
9J", p'~b'blllty of f.i1u,~
100
<,
600
-.......
. <,
o Not peened
500
r-.. l"-
• Peened
-
110 100
/Peened
sn BO
..---- fl--Not peened
400
~-
10
I::
60
50
106 Number of cycles to failure
BO0
50% p,oti.bility 01 f.i1~,e
I 10
100
100
.....
",0
600
<,
500
r-,
400
90
r--
r---- r-rr---- r-----
BO
~
10
I-
60 50
106 Number of cycles
(0
failure
BOO
J", p,lb.b!lIty 011,"u,l 100
600
500
400
0""
.
110
100
.
90
:ii
~~
80
"" f'--.. I'-- tr--o r--.. 0
.........
0
10
. o'
il
60 60
0
106 Number01 cycles to failure
Design stresses. Springs were made from 15.9 to 27.0-mm (% to 1-1/16-in.) diam 8650 and 8660 hot rolled steel and heat treated to between 429 and 444 HB. Springs were shot peened to an average arc height of 0.008 in. on the type C almen strip at 90% visual coverage.
The desirability of conservative design in cyclical service is illustrated in the three charts above, in which the minimum stress used was low. Such data on springs hot wound from bars with as-rolled surfaces are limited, and interpretation is therefore difficult. The value of peening, however, is made quite apparent. Pertinent test data are given above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH. 1978, p 304
487
488
24-5. Coil Springs, HSLA Steels: Effects of Corrosion on Cycles to Failure
~ 540 :<:
Mean stress: 637 MPa (92 ksi)
.; 490 -
-o ::s ~ 440
0&-
0.
e
~
(J)
I
I
I
- 64
r
U
..>:
:r:
0
0
\
C 'tl
0=, •,
:r:
\
48
Cycles to Failure
, , \
H 111
106
0
\
\
~ 50
r
.\.
\
p::
- 57
10 5
104
•, • \ •
52 -
.,-l (/)
(/)
C
o SUP7 _ SUP7-Nb-V
390
(/) (/) Q)
H +J
- 71
ocj-
~
stress: 490 ± 340 t~Pa (71 ± 50 ksi)
- 78
e:tJ-
o SUP7 • SUP7-Nb-V I I I 5x104 10 5 2x10 5 2x10 4 Cycles to Failure
Fatigue life of coil springs: (left) not corroded and (right) corroded, Compositions of HSLA Springs Tested
SUP7 SUP7-Nb-V "
C
Si
Mn
P
S
Cu
Cr
AI
0.58 0.56
2.09 1.94
0.83 0.79
0.014 0.014
0.008 0.008
0.09 0.09
0.14 0.09
0.025 0.021
v
Nb
o
o
0.15
0.18
Fatigue tests on coil springs at a hardness of 50 HRC were performed to examine the feasibility of S UP7-Nb-V to the actual suspension coil springs. When the coil springs were free of corrosion, the result was as shown above (left), in which SUP7-Nb-V has comparable fatigue life to that of SUP7 in any stress amplitude. When the coil springs were corroded, on the other hand, the result was rather different. The corroding condition was as follows: an exposure in a chamber filled with saltwater mist for 1.08 X 104 s (3 h) and a keeping in the atmosphere for 7.56 X 104 s (21 h). After 10and 20 cycles ofthe corroding, the coil springs were loaded with the surface stress of 490 ± 340 MPa (71 ± 50 ksi). The fatigue life of the coil springs subjected to 20 cycles of the corroding are shown above (right). This time, different from the case in the graph at left, there appears a remarkable difference between SUP7 and SUP7-Nb-V. Measurement ofthe surface corrosion depth of the two steels showed no difference.
Source: Toshiro Yamamoto, Ryohei Kobayashi, Toshio Ozone and Mamoru Kurimoto, "Precipitation Strengthened Spring Steel for Automotive Suspensions,"in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 1022
24-6. Leaf Springs, 5160 Steel: Maximum Applied Stress vs Cycles to Failure 2070/ 300 250
U) U)
ILl
lX:
E-<
U) '.-1
1380/200
III
Q::<: ILl <, >-<<1l ...:Il:lo l:lo;:;: c,
«
><
150
. ."'.-... . ...... . . -----....-... b----_ - - a
"
..... "
,.
, •
~
'.
••c
......d
,, -.
690/100
~ ~ u
'.
., f
" 104
-1070 -950 -565 -207
-69 ,
345/5
.
-------
PEAK RESIDUAL STRESS, MPa
105
.. 106
• _ . - ••
+400
107
CYCLES Residual stress and unidirectional bending fatigue data for strain-peened SAE 5160 steel. Applied strain during peening curve a, -0.60%; curve b, -0.30%; curve c, 0% (conventional peening); curve d, preset only; curve e, +0.30%; and curve f, +0.60%.
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; then the S-N curves (see above) were obtained from the same samples by testing in unidirectional bending. In this illustration, 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). This stress-free ASR diagram will be used to predict the endurance limit for the other specimens containing peak stress values of - 1070 (- I55), -950 (- 138), -207 (-30), -69 (- 10), and +400 MPa (+58 ksi). The predicted endurance limit will be compared with the values determined experimentally.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design." in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 82
489
490
24-7. Front Suspension Torsion Bar Springs, 5160H Steel: Distribution of Fatigue Results for Simulated Service Testing 40,-------------------------------....., 25 lots, 300 parts '"
t:
30r--------~A_r/hr7'T_V.l+_-------------____l
~
'0
.8
20
E :l
Z
101------r;..."...,1'A -V/WA-f/.HV/l-V.l1-V'A-t/'J--Y./lH/.,-+-r.,1'A - - - - - - - - _ _ _ _ l
Service life, 1000 cycles
Here are results from simulated service fatigue tests offront suspension torsion bar springs of 5160H steel. Size of hexagonal bar section was 32 mm (1.25 in.), Mean service life, 134,000 cycles; standard deviation, 37,000 cycles; coefficient of variations, 0.28. It must always be considered that results from actual or simulated service testing are likely to vary considerably from results of laboratory testing as shown above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals Park OH, 1978, P 677
24-8. Gears, Carburized Low-Carbon Steel: Relation of Life Factor to Required Life 5.0 4.0
II
~
3.0
u-' 2.0
-
.... .3 u
'"
u.. Q)
...J
1.0
0.5 10 4
106
10 8
Required Life in Cycles
The life factor depends on the required life in cycles. For a single mesh the number of revolutions and the number of cycles are equal. For a gear which has more than one mating member, the life must be equal to the required number of revolutions multiplied by the number of mating gears.
Source: "Bending and Contact Stresses in Hypoid Gear Teeth," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 127
491
492
24-9. Gears, Carburized Low-Carbon Steel: Bending Stress vs Cycles to Failure
(/)lI'l (/)0
LLl ...... 0:::
E-><
(/)
,......
~ '~
6 . 90/ 1. 0
...... 0..
0 ......... Zc;l
LLlo..
a:l::'<:
'-'
3.45/.5
CYCLES Bending fatigue design curves for carburized gears having different amounts of circumferential residual stress at the root-fillet surface,
The bending fatigue design curves for case-carburized gears with the circumferential root-fillet residual stress varying from +138 MPa (20 ksi) to -690 MPa (100 ksi) are shown above. From these curves the residual stress factors at various life cycles were calculated as the ratio of the allowable bending stress for gears with -483 MPa (70 ksi) residual compression to the allowable stress for gears with + 138 (20), 0, -276 (40), and -690 MPa (100 ksi) residual stresses.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 86
24-10. Gears, Carburized Low-Carbon Steel: Effect of Shot Peening on Cycles to Failure
L50 L50 L50
3. 45/ • 5L....::--
'-;-
----'----;;-
----'
105
CYCLES
'" .... 0
1.2
u
« .... Vl Vl
1.1
w
.... '" Vl
...J
« :::>
-
e
1.0
Q
SHOT CLEANED
e
e
~
Vl
w
'"
e
0.9
o. 8 '--.......4- -L..,,----'----;-----..L-:;--.....L.-:;---I 5 7 6 10
10
10
10
10
8
CYCLES Top: Allowable (LSO) bending stress design curves for as-carburized, shot cleaned, and shot peened carburized gears. Bottom: Residual stress factor computed from the upper chart. K a for carburized shot cleaned gears equals 1; that is, the allowable SoN curve for carburized shot cleaned gears is used for design purposes.
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- N curves necessary to calculate the Ka factor. The L50 design curves for as-carburized, shot cleaned, and shot peened gears are shown in the upper chart. The data for as-carburized and shot cleaned gears were obtained on testing six-pitch test pinions on a Four Square Dynamometer. The S-N curve for shot peened gears is derived from the results published by Alman and Black. The lower chart shows the residual stress factors calculated from the S-N curves in the upper chart. It is assumed that the S-N curve for shot cleaned gears is used for design purposes; that is, Ka for shot cleaned gears equals one. 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. The value of Ka deviates from unity with increasing cycles, indicating a more significant effect of residual stress at higher life cycles. At low cycles, the residual stress factor seems to approach one, which means the residual stress has almost no influence on the fatigue properties of a material at high loads. This is obviously because ofthe stress relaxation caused by the cyclic plastic deformation accompanying low-cycle fatigue.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, pp 77, 78
493
494
24-11. Gears, Carburized Low-Carbon Steel: Probability-Stress-Life Design Curves
U'l
o
U) ...... U)
13.80/2.0
L90
Jl.lX
c:<:
E- .......
10.35/1.5
U) 'M
L50
Ul
t:lp..
z .......
'"
6.90/1.0
...... 00..
L10
z ......
Jl.l:'<: !Xl '-'
3. 45/ . 5 L..5 10
...I...-
.1....-
--'
10
8
CYCLES Probability-stress-life design curves for shot cleaned carburized steel gears having a root-fillet surface finish of 5Jl in.; l.e., KJl = 1.
Bending Stresses in Gears In designing gears for a new application, 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. 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. 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. The allowable stress-life diagram characteristic for each material, heat treatment, and surface treatment is usually obtained on testing acceptable commercial quality gears on a dynamometer.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design," in Residual Stress for Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 74
24-12. Gears, 8620H Carburized: Bending or Contact Stress vs Cycles to Fracture or Pitting 500 BENDING
400
(St)
OR CONTACT
GlO~
~}s,
300
(Se)
STRESS, KSI
200 G5~
G90 ~~
_-}st
10 6 10 7 CYCLES TO FRACTURE OR PITTING S-N curves showing the wide difference in cycles to failure between bending and contact stress.
Source; D. H. Breen, "Fundamental Aspects of Gear Strength Requirements," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 66
495
496
24-13. Gears, 8620H Carburized: A Weibull Analysis of Bending Fatigue Data
5,600 .c
..J
c:: - 5,200 Q>
::I
eo
1-c:
4,800
0
c: 0...
4,400 10 5
10 6 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.
Metallurgical data gathered on these gears established the adequacy of the experimental steel (a CRAT steel-computer harmonizing by application tailoring). Although the experimental steel had a significantly lower case hardenability, it quenched out to a 100% martensite plus austenitic structure at the root-fillet surface. Obviously, it had adequate, though not excessive, case hardenability, thus representing an efficient use of alloy hardenability in CRAT steels.
Source: G. H. Walter, "Computer Oriented Gear Steel Design Procedure," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 85
24-14. Gears, 8620H Carburized: T-N Curves for Six-Pinion, Four-Square Tests 8 7 M
0
6
)(
G90 G50 GIO
'" 5
..0
,
c:
4
::::l CI::
aI-
t
BENDING FRACTURE
w
C
----
SPAlLlNG : : : : - - - FATIGUE
~1~~----MIXED
..............
PITTING
2L--
-L..
'-........ '-........G90
1.. . . . .
3
....
'-.....: G50 GIO
----I
-.l
105
CYCLES T-N curves for carburized six-pitch pinion, four-square gear tests.
The above fatigue data show torque versus cycles to breakage, pitting and spalling for a six-pitch pinion test. Note that there is a mixed area where failure can occur from anyone or a mixture of the three modes.
Source: D. H. Breen, "Fundamental Aspects of Gear Strength Requirements," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 66
497
24-15. Hypoid Gears, 8620H Carburized: Minimum Confidence Level; Stress vs Cycles to Rupture
498
C1 Ul 105 0:
....Q.
ai
...J
or
.!!!,.
"
.. .. .. . . ....... .·--_ev .. .. . ..
...
i
M
......
~
<,
<,
.....
.. . <, ~......
.......
a 0:
fill...•
:
1-..........: .....
Iii
105 CYCLES
FOR
"
10 5 RUPTURE - (N)
<,
107
Fatigue life data for hypoid gears. Sloping line indicates minimum confidence level.
Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears." in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 386
24-16. Hypoid, Zero I and Spiral Bevel Gears, 8620H Carburized: S-N Scatter Band and Minimum Confidence Level
-------- ----..........-..........
..........
---- 1;,0--.. 060
o
0
--
---- -- ............ .a"" . ... . .......... ---
co
0
00
0..ll........
<,
<,
~
0
~
<, ..... .; ~"'-"'.
:--..
:
0
0
<, ~~~ <,
....'
.. ... ----... . . .. --. .. .. . . • . . ~ ...... ·8·~
I~
PROPOSED DESIGN LIMIT(NOT OVER 5 PER CENT FAILURES) 1
'l,
.............
" n
n
no
lI) lI)
w
a:::
l-
ll)
105 106 CYCLES FOR RUPTURE - (N) Fatigue data-composite for results obtained by testing various gear designs,
Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 386
499
500
24-17. Spiral Bevel and Zerol Bevel Gears, 8620H Carburized: S-N Scatter Band and Minimum Confidence Level
....,.. ~
«
g.
i.
1'---.",-
d
III
a:
w CL
ai .J
~ I
III
:la:
10
5
.-
<,
.. ..- ..........- .-. .... ... . .'. <, ... . ... . . . . . ...- .. <, -----...... . .... ,... . . . . . . . ... .. 1'--- ...
<;
~
,..,..~"
~
a..
<,
... -.........:
IIII
.........
10 5
C.YCLE:S
10&
fOR
107
RUPTURE -(N)
Graph offatigue life for spiral bevel and Zerol bevel gears. Sloping line indicates the minimum confidence level.
Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 385
501
24-18. Gears, 8620H Case Hardened: Relation of Life Factor to Cycles to Rupture 5 4
~
~
--~
• -
~
3 ~
I 5
5%
2
n0
50%
...
Gl .... :..:i
1'000.
.......
Gl
......
1
u
C Gl
95%
0.9 0.8
-0
c;: C
0
U
0.7
0.6 10J
.....Gl
105
106
Cycles for Rupture
Both strength and durability are fatigue phenomena and therefore display a relationship between stress and life. The life factor for strength may be obtained from the above data.
Source: "Bending and Contact Stresses in Hypoid Gear Teeth," in Source Book on Gear Design, Technology and Performance. Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, p 127
502
24-19. Bevel Gears, Low-Carbon Steel Case Hardened: Relation of Life Factor to Cycles to Rupture for Various Confidence Levels 5
-
4
3
...
~
5%
2
I l>
n
50%
0 u,
...::;
po".
I>
.......
I I
.......
I
, I
1 0.9 0.8
95%
~
I I
l
)] 0.6 103
1O~
106
108
10 7
Cycles for Rupture
The life factor is obtained from the graph above. This depends upon the required life in cycles. For a single mesh the number of revolutions and the number of cycles are the same. For a gear which has more than one mating member the life must be equal to the required number of revolutions multiplied by the number of mating gears. When the required life is less than 6,000,000 cycles on the pinion, the life factors will be different on gear and mating pinion. In cases where the load varies, the designer may wish to determine the equivalent life at maximum torque. One suggested method is as follows: L cp=60L H
T~ [k lnpl+ k n n ( T) 2
5.68
+k)npJ
(Tf; )
5.68
+ ..
+kl/npl/
(
T) i;
5.68 ]
where L cp = required equivalent life in pinion cycles at maximum torque. L H = required total life in hours. k),k 2,k) k,,=proportion of time at torque loads T" Tb T) . . . T"respectively. n PI' n n. n P) n PI/= pinion rpm corresponding to torque loads T" T 2, T) . . . TI/ respectively. T 1, T 2, T) . . . T" = torque loads where T I is maximum torque and T" is minimum torque which will produce a stress above the endurance limit. The required equivalent life in gear cycles at maximum torque may be obtained by multiplying the life in pinion cycles by the gear ratio:
n,
L CG=L cp NG
Source: "Bending Stresses in Bevel Gear Teeth," in Source Book on Gear Design. Technology and Performance. Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, p 149
503
24-20. Gears, AMS 6265: S-N Data for Cut vs Forged 35 .c .....J
= = =-
30
.6.
",,-
,9
25
""
ll...
.8 "C
,~
20
CL CL
"C
co 0 .....J E
15
6 CuI gears • Forged gears
:::I
E ';( co
::;;:
10 5 lQ3
104
10 5
10 6
1Q7
Cycles 10 Failure Fatigue data shown in this chart proved that teeth on precision forged AMS 6265 helicopter pinions have a higher fatigue limit than cut teeth. Loads shown are applied actuator loads. Tooth loads are approximately 33% greater.
Source: "How Gearmaking Methods Compare," in Source Book on Powder Metallurgy. Samuel Bradbury, Ed" American Society for Metals, Metals Park OH, 1979, P 347
504
24-21. Spur Gears, 8620H: S-N Data for Cut vs Forged 120 110 100 "u; a..
90
0 0 0_
80 "0 0 0:::
co 70 e lJ) lJ)
en 60
1% failure
m c
'j§ 0
z:
50 40 30
- - - Cui gears Forged gears
10 5 Cycles 10 Failure S-N curves for spur gears forged and cut from 8620H steel.
As shown above, results of beam fatigue tests indicate that precision flow-forged gear teeth are about 20% higher in fatigue strength than cut teeth.
Source: "How Gearmaking Methods Compare," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 346
24-22. Gears and Pinions: P/M 4600Vvs4615; Weibull Distributions LEGEND ~.
4 5. 6 7 8 3
99._
_5.0
SINTERING TEMP TIME
ALLOY
2~F I MIN
4600V
4600Y
/
2100-F 3MIN
2350·F 3MIN 2350·F 6MIN
4600 V 4600V 4600V 4615
2100-F 6 MIN
/
(BAR STOCK)
80.0
/
::>
400
/
300
3
...J
¢
u.
20.
~
z w
0
a: w Q.
/
/
/
60.0
w a:
/
/
'00 8 60 40
20
/ 1.0 .1
.2
/
/
/
/
/
/
I
/
/
/
/
/
/
--(
/
/
.3
TIME (HR.)
.-
LEGEND ~.
, 2 3
!!::ill
2000 2000 4615
SINTERING
TEMP TIME 2350·F "'6MiN 2350·F 3 MIN (BAR STOCK)
95.0 80.0 600
w a:
40.0
3
300
u.
20.0
¢
~
Z
W
0
a: w Q.
10.0
80 60 40
2.0
1.0
.1
.2
.8
1.0
34~6810
TIME (HR.)
Top: Weibull distribution charts for fatigue testing of actual gears and pinions made from 4600V alloy with various sintering times and temperatures as shown, compared with cut pinions (4615 bar stock). Bottom: Similar to graph at top exceptfor alloy 2000 and 4615bar stock.
Source: P. C. Eloff and L. E. Wilcox, "Fatigue Behavior of Hot Formed Powder Differential Pinions." in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1980, p 308
505
506
24-23. Gears and Pinions: P/M Grades 4600V and 2000 vs 4615; Percent Failure vs Time 99.9
/
LEGEND NO. 950
t
2 3 80.0
1-1
ALLOY 4600V 2000 4615 8AR STOCK 95'/. CONFIDENCE LIMITS
/
40.0
w 0::
/
/
60.0
3
/
/
'i
30.0
~
--l
20.0
lJ...
IZ
w
0
0:: W
(L
10.0 8.0 6.0 4.0
2.0
1.0 .1
2
3
4
5 6
8
10
TIME (HR.)
Fatigue data for actual gears tested in specially designed machines. Presented here are Weibull distributions for the three types of alloys tested.
Since the data from the two powder alloys fell into two groups, it was decided to fit one Weibull curve to all of the data points from each alloy. This was done to obtain more data points for each curve. The results are shown in the graph above, which also graphically indicates the 95%confidence limits on the BID lives. It isplain that the 4600V pinions have superior fatigue life at the stress level of 92,400 psi, and the slope ofthe Weibull curve indicates uniform deoxidation of preforms and therefore less scatter (steeper Weibull slope) in the fatigue data. In the case ofthe 4600V alloy, the sintering temperature should have little effect on deoxidation, since the major alloying constituents, nickel and molybdenum, are readily reducible by CO at temperatures even below 1150 °C (2100 OF).
Source: P. C. Eloff and L. E. Wilcox, "Fatigue Behavior of Hot Formed Powder Differential Pinions." in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1980, P 313
24-24. Gear Steel AMS 6265: Parent Metal vs Electron Beam Welded 2
100
4 68l
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. In the weld evaluations made, excellent mechanical properties were found. Other gear materials tested resulted in compara ble weld joint efficiencies. In general, 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.
Source: N. F. Bratkovich, W. L. Mcintire and Robert E. Purdy. "Electron Beam Welding-Applications and Design Considerations for Aircraft Turbine Engine Gears.t' in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, P 199
507
508
24-25. Gears, 42 CrMo4 (German Specification): S-N Curves for Various Profiles 80
z-,
~. ;.>.
z
>\0
~
40
.........~
.~
30
. .......
~ " ~ ,.... I~
~:~ ! .,..:.-;:::~ "."!.:' ~~~
I
-- _..
3.10 4
.
ll.i
•
J
-
O,I..l,fl
(I ~ (; (I
.::--.
i
~~.
20
--- -- -
:~~.. :. ;. ........
i\:
.- .
. .-
.. --- ..
O.~o2
0, 'lG r\(W/m ~
1.)C G
-J-L
LW--+ SoN curves for various tooth profiles (50% survival probability in the short life range).
The Woehler curves shown above are based (in the sloping section) on a survival probability of 50% at the number of cycles indicated. The horizontal sections of the curves are based on the highest load that can be carried for a minimum life of five million cycles.
Source: H. Winter and M. Hirt, "The Measurement of Actual Strains at Gear Teeth, Influence of Fillet Radius on Stresses and Tooth Strength,"in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed .. American Society for Metals, Metals Park OH, 1980, P 102
24-26. Gears, 42 CrMo4 (German Specification): Endurance Test Results in the Wei bull Distribution Diagram 95
I ,
I I
f/
.....
II ~
II'
I
;-- 70
Of
J
II if
50
~ ...
il 71'
If
30
f ,I:
20
I
J~!'I
10 8
A -1·-
,
l-
-- _..
----f I - - 1-I
6 4
~
L
3.,0'< 4
- --.-
6
8 1.'0:>
. .
[
2
3
t. 5
,-W
(;
. -
8
1.1
--~
Endurance test results in the Weibull distribution diagram.
The parameters ')I,A are adjusted to the test points. An example is shown in the graph above.which represents in such a probability grid the test results for gears of one tooth form. Scale of ordinates is the failure probability A = 1- W= if (n + I) for test i out of n test results sorted to the number ofload cycles at which fracture occurred. The test points are approximated with a straight line. From this curve we are able to read, respectively, life values L IO , L so, and LJo for 10,50, and 90 percent failure probability, or 90, 50 and 10 percent survival probability. A more adequate approximation by the theoretical distribution is achieved by a three-parameter Weibull distribution. This formulation produces a minimum endurance Lo, which is reached by all test pieces. Also, the above chart shows the compensating curve which results from the formulation of the three-parameter Weibull distribution.
Source: H. Winter and M. Hirt, "The Measurement of Actual Strains at Gear Teeth, Influence of Fillet Radius on Stresses and Tooth Strength,"in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed.. American Society for Metals, Metals Park OH, 1980, P 102
509
510
24-27. Bolts, 1040 and 4037 Steels: Maximum Bending Stress vs Number of Stress Cycles 100 0
0
•
1040 steel I
-
100
-
95
• 4037 steel
615
s:
.
:::iE
~g'
0
650
"v;
"'". ~
625
-
0.0
:0
~ )(
~
•
600
.n .
~
Cl
c 90 :0 c
.
..
.0 )(
:::iE
~o.
.
of>
~-
~~-
-
85
515 o
0
•
•
0
>--1-
Number of stress cycles
The bolts (7'8 by 2 in., 16 threads to the inch) had a hardness of 35 HRC. Tensile properties of the 1040 steel at three-thread exposure were: yield strength, 1060 MPa (154 ksi); tensile strength (axial), 1200 MPa (175 ksi); tensile strength (wedge), 1190 MPa (173 ksi). For the 4037 steel: yield strength, 1110 MPa (161 ksi); tensile strength (axial), 1250 MPa (182 ksi); tensile strength (wedge), 1250 MPa (182 ksi).
In general, if bolts made of two different steels have equivalent hardnesses throughout identical sections, their fatigue strengths will be similar (see above S- N data), 'as long as other factors such as mean stress, stress range, and surface condition are the same. If the results of fatigue tests on standard test specimens were interpreted literally, high-carbon steels would be selected for bolts. Actually, steels of high carbon content (more than 0.55% carbon) are unsuitable because they are notch sensitive. The principal design feature of a bolt is the threaded section, which establishes a notch pattern inherent in the part because of its design. The form of the threads, plus any mechanical or metallurgical condition that also creates a surface notch, is much more important than steel composition in determining the fatigue resistance of a particular lot of bolts.
Source: ASM Committee on Carbon and Alloy Steels. "Threaded Steel Fasteners." in Quality Control Source Book. A. K. Hingwe, Ed .. American Society for Metals, Metals Park OH, 1982. P 206
24-28. Bolts: S-N Data for Roll Threading Before and After Heat Treatment 450 r - - - , - - - - , - - - - - , - - - , Roll threaded before heat treatment
60
375 1 - - - t - - - - - 1 - - - - t - - - - - j 50 3001---t-----1----t-----j 40 225
~
30 U5 150 f-----11-
20
751----t------I---'
o
L-
! -_ _-'----_ _
10
~_~
104 450 , - - - , - - - - , - - - - - . - - - , 60 375 -
50
300 40
'" :;;:
0-
::i 225 I - - - t - -
1" U5
150 f - - - - t - - - - t - - - + - - - - - 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. Upper graph represents four different lots of bolts that were roll threaded, then heat treated to average hardness of 22.7, 26.6, 27.6, and 32.6 HRC. Lower graph represents five different lots that were heat treated to average hardnesses of 23.3, 27.4, 29.6, 31.7, and 33.0 HRC, then roll threaded. Bolts having higher hardnesses in each category had higher fatigue strengths.
Other factors being equal, a bolt with threads properly rolled after heat treatment-that is, free from mechanical imperfections-has a higher fatigue limit than one with cut threads. This is true for any strength category. The cold work of rolling increases the strength at the weakest section (the thread root) and imparts residual compressive stresses, similar to those imparted by shot peening.
Source: ASM Committee on Carbon and Alloy Steels, "Threaded Steel Fasteners," in Quality Control Source Book, A. K. Hingwe, Ed., American Society for Metals, Metals Park OH, 1982, P 202
511
512
24-29. Power Shafts, AMS 6382 and AMS 6260: Electron Beam Welded vs Silver Brazed Joints .. ELECTRON BEAM WELDED JOINTS
• 60
o o
TEST TEMPERATURE, 500-600 oF
.
Q
x
SILVER BRAZED JOINTS
50
III
o
S/N-2
Z :l :I:
u
..
40
z +1 w
5
30
0:
o
I-
>0: o
!;i
20
0:
m
>
-
--
r-- l I-
.. ----
~~ S/N-2
SIN-I ROOM TEMP.
i----
r---r--.•
..
S/N-3
o0..
....
SIN-I
:---
--. "-
-I---
10
CYCLES TO FAILURE
S-N curves for electron beam welded versus silver brazed power shafts made from AMS 6382 and 6260 alloy steels.
In the welded shafts, failures occurred apart from the weld, while in the brazed units all failures occurred in the brazed joints.
Source: S. M. Silverstein, V. Strautman and V. R. Freeman, "Application of Electron Beam Welding to Rotating Gas Turbine Components," in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, p 187
24-30. Axle Shafts, 1046, 1541 and 50854 Steels: S-N Data for Induction Hardening vs Through Hardening
513
:\, B & C
I1'I1lUCf IO~ HARDENED 1541 1046 E - THROUQI HARDENED 50B54
D-
~on if.
'"" if.
v: 1 SO
?i
.
..9."Il!':.? ....
,.....
-,
tr.
".
.... ..........
:~
\
...
~
"
.....
u:: ] on
,
>c::: < ~
0;
., ............ --
\',
.......
-. -'-."
·'-'-'-B
......... ::::.:::-:.:::-::.: ~ "
50
-
APPLIED ~ STRESS GRWIE~T 200
400 600 DISTA.'JCE FRO~l SURFACES (.001")
A, B & C - rxructrox HARDE\'ED 1541 D 1046 E - TI-JROllGH HARDE\ED 50B46
..... 140
IF. ~
u: 100 u:
..... .......~ . " .. .~ '
n
':-.."
2
l-
i/;
~
800
60
.,
-, . "
f:S
u: 20 10 2
]0 3
10 4
"
:"::..,,:,,'~ c .'·-._.-E ----B lOS
10 6
NO. OF CYCLES TO FAILURE Top: Axle-shaft strength gradients in terms of shear yield strength. Bottom: Fatigue performances of axle shafts as a function of strength gradient.
Induction hardened 0.40% carbon steel axle shafts were developed to replace through hardened alloy steel shafts for both product- and cost-improvement purposes. This was accomplished after a rather comprehensive bench-test program, which examined variables such as surface hardness, core hardness, gradient strength, distortion, composition, and surface-condition effects. The more promising approach was then subjected to chassis, proving-ground, and in-service testing. Some interesting reflections can result from examining some of the fatigue data that were generated. That the through hardened concept was vulnerable can be surmised by considering the stresses developed in a full-floating splined shaft loaded in torsion. The stresses are a maximum at the surface and drop linearly to zero at the center. At the spline, the stresses drop more rapidly at the onset due to the stress concentration caused by the spline. The upper graph shows the stress gradient in the body area when the shaft is loaded to 110,000-psi shear stress. Also plotted on this graph are the shear yield-strength gradients (converted from hardness) of the production alloy shaft, along with several experimental induction hardened shafts. One would expect the through hardened shaft to have a surface-origin failure and to be lower in strength, since its surface is the lowest hardness. Also, the high strength of the center ofthe shaft is essentially wasted, since it is lowly stressed. Gradient strengthening by induction hardening provides a means of providing a better strength match for the stress gradient. The lower graph gives the fatigue curves established for shafts having the strength gradients shown in the upper graph.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure, American Society for Metals. Metals Park OH. 1979, P 88
514
24-31. Steel Rollers, 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 99·9
SAE 8b2011 - Reheat
99·0
Carbo of
95'0
0 Grou 0 II Group Q
90·0
• Group
80·0
----
70'0
R
1750 1800 1900'
Quench
.
Sliding
..
~I
21 21
Slope
Cor r .
Coef.
B
.89 .97 .98
2.199 I. 463 2.961
·vacuum
------
AJ·l
60·0 50·0 40'0 -
Data corrected to
= 400 ksi Sc___ ____ A= 0.5
30'0
20·0 ~
.
a: :>
:;;
.... 10'0 I-
Z ~
U
a: ~
e,
5'0 4'0 3·0
"0
R
0
Q 1'0
0'5 0'4 0'3 0·2
105 0"
I
106 4
5
6 7 8 9 ,
8 10
7 2
3
c v c i es
4
5
6 7 • 9 ,
2
4
5
6 7,19
S-N data, Weibull probability plot: Effect of three carburizing temperatures on surface fa-
tigue for carburized 8620" steel. All were slow cooled and reheated for quenching. This technique improved fatigue characteristics compared with direct-quenched rollers.
Source: S. L Rice. "Pilling Resistance of Some High TemperatureCarburized Cases," in Source Book on Gear Design. Technology and Performance. Maurice A, H, Howes, Ed" American Society for Metals, Metals Park OH, 1980, p 234
24-32. Steel Rollers, 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 99 .•
=...,
1-
.'·0
SAE 8620H - Direct Quenched Carbo
OF 9"0 '0' 0 1-0=--Gc..r_o_u.... p_N
Sliding \
Corr.
Coef.
r-
21 .90 1750 +f. ._ _r----1 .97 21 1800 .97 21 1900' 10.0 ~.u!!J"--"S'---'-"=--=----'-".:.-----=:..:..:.:...:..,~J_------+f ...- r _ - - - - _ j
o
70.0
Group P
~a"'cull ...umlll-.
_f_.=_"~-----_c,.-_r'--:_---__1
6&:~(--------------_:_ . .+-------__,f.;;;;:"M,./_------_I 'O·ol---------------.'f--------r..-f---i~-------__I 4o·01---------------;.t----------tl~-'":f_--------__I
Data Corrected to 30·01--1>:c-~W9__1
c
A = 0,5 20·0
. JO·Oj----------r--='---------if---------------_f
~
I-
z
U 0:
.. "0 - - - - - - - - f - : ; . ; o - - - - - - - - - - T - . H........---------------1
"0 t - - - - - r - - - - - - - - + - j ' - - - - - - - - - - - - - - - - - - - _ _ I N
S
P
'-0
D"
0" D"
0·3
0'\
,105
345678"
2
3
C 't'
c i es
4567."
45.71.
S-N data, Weibull probability factor: Effect ofthe three carburizing temperatures on surface fatigue for carburized 8620H steel. All were direct quenched from the temperatures shown above.
Source: S. L. Rice, "Pitting Resistance of Some High Temperature Carburized Cases," in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed.. American Society for Metals, Metals Park OH, 1980, P 233
515
516
24-33. Linkage Arm, Cast Low-Carbon Steel: Starting Crack Size vs Cycles to Failure
III OJ
s: u c .300
E E I
7 0
.250
0
W
6
N
en :.:: <{
i=
3
a::
en
<{
4
.150
:.::
U
af; 0.63 in. 116 m rn )
a:: u
;:!
iii 5
U
<.9 Z
N
(136MPa) .200
W
a:: U
<.9 Z
i= a::
.100
<{
2
I-
en
.050
1000
10,000
CYCLES TO
100,000
FAILURE. N f -
Fatigue life of a linkage arm as a function of starting crack size.
The variation of the fatigue life, NJ, with the starting flaw size a.; is shown in the diagram above. The fatigue life increases dramatically at very small a i values. The far curve shows that in the long life regime the final crack size has only a small effect on Np This is because fatigue crack growth rates are very low at low ~Kvalues and hence the greatest fraction offatigue life is spent at the smaller crack sizes. Since the controlling parameter is ~K, low life for small crack sizes is possible at high cyclic stresses. The second set of curves shows that doubling the cyclic stress range reduces the fatigue life by about an order of magnitude. Also, if the starting ~K value is high, the final crack size has a larger effect on the cyclic life. The above diagram shows the importance of adjusting both the cyclic stress and starting flaw size to optimize the fatigue life.
Source: Steel Castings Handbook. 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH. 1980, P 4-17
24-34. Notched Links, Hot Rolled Low-Carbon Steel: S-N Data for Component Test Model ... - Component Tests
I. 0 r - - - - - - - - - - - . : . . . - - - - - - - - - - - ,
~:
~.., _
......
0.7
~~notched
~~~ ~/2Z7!?2Z .~
0.6
_./Coml
---------_--=--~~~nent Tests
0.5
<, .l ~~ Local-Strain Model
--A
rr .............& ~1 ......
04 .
0.3
S-N / Model
R=-1
~@;/?/!?!Z2Z
I
0.2L..--------1------.........L~------J 7
~
loll lOll Fatigue Life. Nj • cycles
10·
COMPONENT TEST MODEL. The most straightforward life prediction model of a component is developed from fatigue tests of the component itself. The component is cyclically loaded in a manner that attempts to simulate service and the model is the plot of the test results. The cyclic load or nominal net-section stress is plotted versus cycles to crack initiation N; or to failure Np Example: Component fatigue tests were performed on the notched link ofthe previous examples; the results are listed in the table and shown graphically above. The fatigue strength for crack initiation at one million cycles is: iJ.S = 0.47 Sli
106 cycles The three models-S-N, local strain, and component tests-are compared above, and the three predictions for fatigue strength (iJ.S) at one million cycles are, respectively, 28, 42, and 47% of the tensile strength Sli' The local strain prediction is closer to the behavior observed in the component test than the S-N prediction. Component Test Results, Hot-Rolled Carbon Steel Notched Link Tensile strength,
Yield strength,
Cycles to crack initiation,
Su,MPa
S}",MPa
N;
417 388 366 366 402 417 366 402 388
242 236 239 239 235 242 239 235 236
0.818 0.880 0.936 0.690 0.628 0.604 0.444 0.404 0.418
1.36 X 1.69 X 2.62 X 6.99 X 8.24 X 1.05 X 1.64 X 2.21 X 2.99 X
10' 10' 10' 10' 10' 105 10" 10" 106
Source: Harold S. Reemsnyder, "Constant Amplitude Fatigue Life Assessment Models," in Proceedings of the SAE Fatigue Conference P-I09, Society of Automotive Engineers. Inc .. Warrendale PA, 1982. P 127
517
518
24-35. Fuselage Brace, Ti-6AI-6V-2Sn: Fatigue Endurance of HIP-Consolidated Powder
.;;; .:.!
tl' 120 ~
:;; E ::> E 100
'x
'"
~
80
60
Annealed plate (min)
-..; 10'
10'
m 10·
10'
Number of cycles
Fatigue endurance of Ti-6AI-6V-2Sn powder consolidated by HIP at 1650 of (900°C). Significance of boxed numbers is as follows: 1: HIP run #1, as machined. 2: HIP run #2, vacuum annealed for 2 h at 1300 of. 3: HIP run #2, vacuum annealed for 16 h at 1300 ° F. 4: HIP run #2, vacuum annealed for 24 h at 1300 OF. 5: HIP run #4, vacuum annealed for 24 hat 1300 OF.
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. The tensile and toughness properties developed compare well with the average values for forgings. The fatigue endurance limit of H lP powder developed in this program is given in the SoN diagram above. Here the HIP data points lie within the band for annealed forgings and plate.
Source: J. H. Moll. V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications," in Powder MetallurgyApplications. Advantages and Limitations. Erhard Klar, Ed.. American Society for Metals. Metals Park OH, 1983, P 288
