| This program investigated the effects of varying alternating shear stresses and static tensile and compressive mean stresses applied normal to both maximum shear planes on the fatigue life of a tubular specimen. Once the behavior of static mean stresses normal to the maximum shear planes was established, the assumption was made that the treatment of residual stresses in multiaxial fatigue was identical to that of static mean stresses. A methodology for predicting the fatigue lives of induction hardened shafts was developed and was tested.; Tubular specimens of SAE 1045 steel, BHN 456 and 203 were tested in axial tension-compression with alternating internal/external pressure. For a given fatigue life, an approximately linear relationship was found between the applied normal static mean stresses and the maximum cyclic shear stresses on the critical shear planes. For a constant shear stress, the fatigue life remained constant with increasing mean stress for static tensile mean stresses larger than 500 MPa and 76 MPa, (Sint), for the hard and soft steel, respectively.; To test the hypothesis that residual stresses had the same effect as static mean stresses, 13 induction hardened shafts were tested in pure torsion. The residual stress and hardness distributions were determined in the shafts by x-ray diffraction and a series of hardness tests, respectively. The shear strain distribution through the shafts was determined at the potential failure locations. For each potential failure location and radial depth, a Modified Findley parameter value was calculated and the hardness level was recorded. This value was compared to the parameter versus fatigue life curve from the tubular specimen results for the appropriate hardness level. This methodology allowed the crack location along the length of the shaft as well as the depth of the crack initiation site to be calculated. These predictions were verified with the experimentally obtained results. |
