It is well known in the literature that fatigue failures originate from the weakest element within the component risk-volume (i.e., the region of material subjected to a stress amplitude close the maximum applied stress). In particular, for metallic materials the fatigue crack typically originates from the most critical surface defect in the High Cycle Fatigue (HCF) region; whereas, it generally originates from the most critical internal defect in the Very High Cycle Fatigue (VHCF), at stress amplitudes below the so-called ‘transition stress amplitude’. Since the fatigue crack originates from critical defects, the P-S-N curve must necessarily take into account the random distribution of critical defect size. However, in the literature, there are few statistical models capable of modelling the dependency between the fatigue life and the defect size or taking into account the presence of different failure mechanisms (e.g., surface crack nucleation in HCF and internal crack nucleation with fish-eye morphology in VHCF region).
In the present paper, a new software for the estimation of the P-S-N curves is proposed. The statistical P-S-N models that have been recently proposed by the Authors are considered in the software and described in the paper. The procedure for the parameter estimation is also explained in detail: both the least square method and the maximum likelihood principle are implemented, depending on the presence of runout specimens. The software is finally validated with simulated datasets and with experimental datasets taken by the Authors or available in the literature.