The fatigue performance under service load of transmission shafts and spline couplings significantly influences their reliability. They are typically loaded by multiaxial stress states owing to the superposition of torsion loading and rotating bending, that make fatigue lifetime prediction challenging. This thesis was motivated by an industrial issue faced in practice reported by Dumarey Automotive Italia S.p.A., part of the international Dumarey Automotive Group, where commercial fatigue software Fe-Safe predicted anomalously low safety against fatigue in the middle of a spline groove. In order to decide whether such an anomaly is caused by a physical phenomenon or by numerical artefacts within invariant-based formulations (e.g., Von Mises), a postprocessing tool independent from the Fe-Safe tool was developed in Python.