PSI - Issue 39

Daniela Scorza et al. / Procedia Structural Integrity 39 (2022) 503–508 Author name / Structural Integrity Procedia 00 (2019) 000–000

505

3

In view of Eq.(1), Yanase and Endo (2014) proposed an analogous equation to estimate fatigue limit under torsion according to the two above assumptions: ( ) ( ) 1 6 1 19 120 wl max . HV area τ + = (2)

The fatigue limits of Eqs. (1) and (2) are assumed in correspondence of ( ) 7 1 10 ⋅ loading cycles.

3. Experimental campaign examined 3.1. Multiaxial fatigue test data

The examined experimental campaign was carried out by Machado et al. (2020) and refers to specimens extracted from a broken power generator crankshaft made of AISI 4140 steel. The main mechanical properties are: 689 y MPa σ = , 900 u MPa σ = , 20% u ε = and 2 320 f HV kg mm = . The tested specimens have the geometry shown in Figure 1, and they were subjected to uniaxial and multiaxial fatigue tests under load control, according to the international standard ASTM E466-15.

Fig. 1. Specimen geometry with sizes in mm (Machado et al., 2020).

The values of the ratio between shear stress amplitude xy ,a τ and normal stress amplitude x ,a σ were: (a) 0.0, 0.5, 1.0, 2.0 and ∞ with a phase shift β =0°, and (b) 0.5, 1.0 and 2.0 with a phase shift 90 β = ° . The specimen complete rupture was used as the failure criterion, whereas the run-out condition was in correspondence of ( ) 6 2 10 ⋅ loading cycles. Further details may be found in Machado et al. (2020). 3.2. Estimation of the fatigue strengths through the area -parameter model In order to compute wl σ (Eq.(1)) and wl τ (Eq.(2)), an inclusion content analysis was performed by Machado et al. (2020), and the main steps are hereafter summarised: (i) two specimen cross-sections were examined by a Scanning Electron Microscope: one perpendicular to and one forming 45° with the specimen longitudinal axis; (ii) for each inspection area, the biggest inclusion was identified and its max,i area computed, by repeating such an operation 60 times; (iii) the statistics of extreme value theory was used to estimate the largest inclusion size max area in both cross sections. In order to compute the max area , two volumes need to be set: the control volume V , that is the volume associated to the useful cross-section and the standard inspection volume 0 V . The first is arbitrarily chosen, whereas 0 V is constant and depends on the size of the pictures taken by the SEM (that is, 3 0 00617 . mm and 3 0 00643 . mm for the cross-sections perpendicular to and forming 45° with the specimen longitudinal axis, respectively). In the present