PSI - Issue 57

Jan Papuga et al. / Procedia Structural Integrity 57 (2024) 79–86 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

84 6

along an interesting finding that the response of the pair DV & CROSS is similar. In the case of in-phase loading, it seems that the position of both curves drawn there are equidistant, while in the case of out-of-phase loading, they reach nearly identical results. Apparently, the difference in the response to the phase shift documented in Fig. 3 is just enough to bring the original mild difference in the in-phase regime to overlapping in the response to OP configuration.

Fig. 3. Response of various criteria to the phase shift normalized in the FI parameter by the response to the in-phase load regime.

Fig. 4. Fatigue index errors obtained for both types of load regimes for the four evaluated criteria.

The DV solution provided the best output for the in-phase case, but the too dramatic phase shift effect decreases its overall fatigue strength estimation quality. In the case of LZ criterion, we see the worst estimates in the IP configuration, reaching quite non-conservative results there, in the HCF region above 100,000 cycles above all. Its quality is much better in the OP case, where it reaches values alike to QCP. QCP is the optimum solution from the evaluated four criteria, reaching the steadiest response across both load regimes. Obvious is however that the worst quality of the fatigue strength estimation is achieved for the highest lifetimes. There, DV and CROSS are too non conservative, while the difference between estimation quality in IP and OP cases is the highest for LZ and QCP. 4. Discussion Based on the expected responses from Fig. 3, OP loading should generate substantially lower equivalent stress amplitude than the IP case, and thus it should lead to longer fatigue lives for the same stress amplitudes in the case of