PSI - Issue 57

Lucas Carneiro Araujo et al. / Procedia Structural Integrity 57 (2024) 144–151

149

6

Author name / Structural Integrity Procedia 00 (2019) 000 – 000 Table 2. Fatigue strengths of AISI 4140 steel from √ parameter model. Defect Fatigue strength (MPa) Non-metallicinclusions 271 Micro-hole 220

4.2. Fatigue data and predictions of the proposed model (SWT mod )

The proposed multiaxial criterion (SWT mod ) generates a fatigue strength curve that separates the safe region from the failure region. Stresses below this curve are expected to result in run-outs, while stresses above it indicate potential failure. By considering the fatigue limits obtained from the √ parameter model, the proposed model is expected to account for the influence of small defects on fatigue strength. Figures 2 and 3 illustrate the comparison between experimental data points and the curves obtained with the proposed model for the fatigue strength for non-metallic inclusions and superficial micro holes, respectively. Dashed lines represent error bands of 15%. Full symbols represent specimens that failed, while empty symbols represent run-outs. Some symbols may overlap, indicating stress levels where both failures and run-outs occurred. Fig. 2 specifically depicts data from smooth specimens, with Fig. 2(a) showing in-phase loading conditions and Fig. 2(b) presenting out-of-phase loading conditions.

= 0

/

= 0

/

( )

(a)

Fig. 2. SWT mod prediction and experimental data of smooth specimens, (a) uniaxial and combined in-phase and (b) combined out-of-phase data. Additionally, Fig. 3 presents the experimental data obtained from specimens containing the superficial micro hole. Fig. 3(a) displays the results for uniaxial and combined in-phase loading conditions, while Fig. 3(b) represents the combined out-of-phase loading conditions. Remarkably, the predictions turned out to be very accurate, with the vast majority of points within the 15% error bands. There was a slightly conservative trend for the smooth specimens, with some run-outs in the fault region, Fig. 2(a), and a slightly non-conservative trend for the micro-hole specimens, with some faults in the region where run outs were expected, Fig. 3(a). Except for the uniaxial tensile-compression test data for smooth specimens, all points were within or very close to the 15% error bands. The accuracy of the criterion was also assessed using the error index ( ) in Eq. 10, which quantifies the relative difference between the experimental result ( ) and the predicted result ( ). The average and maximum values of the errors obtained from the multiaxial criterion, considering various loading conditions and specimen conditions, are presented in Tab. 3.