Observations of the orientation of crack formation and growth in wrought metals have been the basis for many shear strain-based low-cycle multiaxial critical plane fatigue damage models. One common factor with most of these models is that they consider a maximum shear amplitude often modified by a stress or strain, or both, normal to the maximum shear orientation. Considering the variety of multiaxial fatigue loadings, one factor that can drastically influence fatigue damage is mean stress/strain. Critical shear strain-based approaches using only modifying normal strain terms cannot account for the decreased damage observed when mean stresses relax, as they often do for constant amplitude low-cycle loadings. An alternate fatigue damage criterion, employing the maximum stress normal to the shear orientation, was recently proposed to account for this effect.
The importance of mean stresses/strains has been demonstrated with uniaxial tests. To verify the validity of the proposed criterion, twelve multiaxial loadings with mean stress, mean strain, or both are investigated employing combinations of axial-torsion and biaxial-tension. Surface crack orientation and fracture surface features are used to demonstrate that the terms employed in the mathematical formulation of the damage parameter are appropriate for the fatigue characterization of SAE 1045 steel and Inconel 718 at room temperature. Correlation of the physical fatigue damage mechanism with the modeling parameter terms results in increased confidence when attempting to predict the fatigue lives for other stress-strain states.