PSI - Issue 77

Bastian Roidl et al. / Procedia Structural Integrity 77 (2026) 119–126 Bastian Roidl / Structural Integrity Procedia 00 (2025) 000 – 000

122

4

The Kohout & Věchet (K&V) model was chosen because it effectively captures the fatigue be havior across the entire lifespan — from low-cycle fatigue (single-digit cycles) to the transition into the fatigue limit region (more than 10 6 cycles). Unlike the conventional power-law model, the K&V approach provides a more accurate description of this full fatigue spectrum. The fatigue curves themselves (Fig. 2) are not the primary focus of this paper and will not be analyzed in depth. The curves were used to determine the fatigue strengths for specific numbers of cycles for each series. Matusu et al. published data of a fatigue curve obtained under comparable experimental conditions and with specimens shaped like geometry A [14], [15]. Comparing the fatigue curve of Geometry A with their data, no significant discrepancies can be found. Data from tensile tests, which have also been published, are at a similar level [14], [15].

Fig. 2 (Left) Fatigue curve of A, G and H geometry with load ratio of R = 0.1. (Right) Fatigue curve of A, G and H geometry with load ratio equal to R = -1. Fatigue curves are fitted with K&V regression. A-type specimens have cross-sectional area at the critical cross-section of 64mm², G of 30mm² and H of 10mm². 2.4. Mean stress effect and effect of geometry By testing specimens with load ratio R =-1, no mean stress is applied. The specimens are only loaded with the amplitude under fully compression and tension during one cycle. In contrast, at R =0.1, the specimen is permanently under tension because the mean stress exceeds the amplitude. Under fully reversed loading ( R =-1), all specimen geometries exhibit higher fatigue strength across the entire investigated lifetime domain (see Table 2). At a positive load ratio ( R =0.1) the curves of variants G and H converge, yielding a minimal difference in fatigue strength (Fig. 2, left). Different behavior is revealed in experiments under a load ratio of R =-1. This can be attributed to the elevated maximum stress experienced under R =0.1 compared to R =-1, which strongly affects crack propagation over short crack paths. Despite possessing the largest cross section – and thus an expected size effect – A-specimens achieve the highest fatigue strength at R =0.1. Under fully reversed loading, however, G-specimens, with their small critical cross section, display the lowest fatigue strength, presumably due to the immediate impact of crack initiation initiated in the small section. The fatigue responses of H- and A- specimens under R =-1 are very similar, suggesting that the size effect during crack formation in the H-geometry shifts its S-N curve toward that of variant A. The mean stress sensitivity is a factor that shows how much influence mean stress has on the fatigue performance. To compute it, both results of both load ratios have to be used: = ( =−1 ) − ( =0.1 ) ( =0.1 ) (2)