PSI - Issue 68

Giuseppe Macoretta et al. / Procedia Structural Integrity 68 (2025) 974–980 G. Macoretta et al. / Structural Integrity Procedia 00 (2025) 000–000

978

5

3.2. S-N curves and notch sensitivity factor evaluation The fatigue data are presented through S-N diagrams in Fig. 5, where the fitting lines correspond to a 50% probability of failure. Both the smooth and notched specimens exhibit a fatigue behavior characterized by the occurrence of a fatigue limit before 10 5 cycles. The S-N curves feature a knee point in the range 5∙10 4 ÷ 8∙10 4 cycles. Beyond these values, the scatter of the data becomes marked, and most of the specimens tested at slightly lower load values did not fail after 1∙10 7 cycles. The fitting lines in Fig. 5 were calculated considering only the specimen broken within the region between 10 3 and 10 5 cycles. The slope of the S-N curve of the smooth specimens was found to be slightly greater than the one of the V-notched specimens. The inverse slope of the fitting line resulted to be directly proportional to the stress concentration factor, passing from 2.4 for the smooth specimens to 3.9 for the k t 3.5 ones. When calculated at 6∙10 4 cycles, the fatigue notch sensitivity, q , resulted to be 0.13 for the blunt notches ( k t 1.5 and 2) and 0.22 for the sharper one, as listed in Table 1. The low notch sensitivity for the blunt notches is due to the surface roughness contribution can hide the effect of the macroscopic notch, as it reduces the performances of the smooth and notched specimens in an analogous way. The values are lower than what has been observed on similar notches produced by L-PBF and tested at room temperature [17], but the same trend between the notch radius and the notch sensitivity factor has been observed. As shown in Fig. 4, the effective notch produced by the L-PBF process in the case of the k t 3.5 notch presented a markedly lower radius (0.13mm) than the nominal one (0.22mm), which was implemented in the FE model. It can explain the reason for the increased notch sensitivity for these specimens. As shown in Fig. 6, the machine operating frequency shows an abrupt drop for both the smooth and notched specimens. This behavior was observed independently from the nominal stress values. Until the last 3% to 5% of the fatigue life, the machine frequency remains constant or features a slight increase, demonstrating that crack propagation took a negligible portion of the specimen fatigue life.

Fig. 5. S-N curves of the smooth and notched specimens.

Table 1. Fatigue notch factor and notch sensitivity values.

k t 1.5

k t 2

k t 3.5

k f

1.06 0.13

1.14 0.13

1.55 0.22

q