Static Axial Tension or Compression Stress Effects on Shear Cracks Initiating under Cyclic Torsion Loads in Multiaxial Fatigue Tests

•static axial compression loads may be detrimental in cyclic torsion test lives.•shear-based damage models using deviatoric components can quantify such effects.•peak correction terms are only applied to the elastic component of the shear ranges.•validation from cyclic torsion fatigue tests with superimposed axial stresses.•shear-based models should only be calibrated from torsional shear-driven crack data. This work proposes models that can better fit and explain apparently contradictory mean stress effects when superimposing axial tension or compression loads on cyclic torsion fatigue tests. It introduces new shear-based multiaxial fatigue damage models, modifying the Fatemi-Socie equation to consider the peak deviatoric stress normal to the initiating shear crack as the driving force for the peak or mean effects of the normal stresses. This deviatoric hypothesis can quantify the odd detrimental effects of static axial compression observed on axial shear cracks tested under cyclic torsion. Additionally, applying the peak deviatoric correction term only to the elastic component of the shear strain range allows a better description of the increasing mean normal stress effects in high-cycle fatigue, eliminating the need for a variable normal stress sensitivity coefficient in the multiaxial fatigue damage parameter. To evaluate the proposed shear-based models, their life and crack initiation direction predictions are compared for cyclic torsion experiments both with and without superimposed static axial stresses for steels and aluminum alloys, including both literature and original data. The experiments show that axial compression can be detrimental to the fatigue initiation life, depending on the cracking mode and presence of bifurcations, which effects could be explained from the proposed deviatoric damage parameters.