PSI - Issue 75

Ralf Glienke et al. / Procedia Structural Integrity 75 (2025) 474–488 Ralf Glienke et al / Structural Integrity Procedia 00 (2019) 000 – 000

482

9

For this purpose, two further series were carried out with comparable SCFs, with the yield strength as maximum stress  max = f y (#04) or with high mean stresses 0.44 ≤ R ≤ 0.5 (#03). Based on this, the S-N curves become steeper, but still show a significant improvement in fatigue strength compared to the reference DC 90. 3.5.2. Findings on thermal cut edges The tests on specimens with laser cut edges made of S235JR (#07), which represents the lowest material strength according to IIW Recommendations (2024) or FprEN 1993-1-9 (2024), were carried out appropriately at R = -1.0. According to Schütz (1965), there is almost no mean stress dependence for mild structural steels. The additional blast cleaning treatment improves the fatigue strength by a factor of k Imp = 1.10 and flattens the S-N curve for test series #08. A shift in the crack location due to the blast-cleaning treatment was not observed for this series. Decisive for the failure was not the hardened edge from laser cutting but the top or bottom of the specimen. In the case of specimens with oxy-fuel cut edges (#09), with greater expansion of the heat-affected zone and a higher surface roughness due to the burning marks, the blast-cleaning (#10) leads to a significant fatigue strength improvement (k Imp = 1.31) and flattening of the S-N curve (m BC / m = 3.5). A shift of the crack initiation location did not occur due to the blast cleaning. 3.5.3. Findings on base material The base material specimens in the as-rolled condition were tested at R = -1.0 to determine the material fatigue strength at zero mean stress. The FKM approach was used to evaluate the results, whereby the amplitude of the material fatigue limit at zero mean stress was converted to the stress range at N D = 2∙10 6 . Considering the surface roughness reduction for the mill scale according to the FKM-Guideline (2020) and an assumed slope parameter of the S-N curve in the finite life region for notch-free specimen of k  = 15, the test data fit very well for S235JR (#11) and S355J2+N (#13). The blast-cleaning treatment again improves the fatigue strength of all series (#12, #14 and #16), resulting in k Imp = 1.03 … 1.15. The FKM-Guideline (2020) specifies a surface treatment factor for shot peening of K V = 1.1 … 1.2 on the fatigue endurance limit for notch-free components, which roughly corresponds to the test results for blast-cleaning. The results for the S690 are remarkable. Although the fracture events for the specimens with mill scale (#15) are located above the synthetic S-N curve, the test-related slope is significantly steeper compared to the specimens of the mild steel grades. The blast-cleaning treatment Sa 3 ensures an improvement in mean number of load cycles, but the scatter increases significantly. This could indicate that the blast-cleaning treatment is not as effective for the high-strength materials, as these are also more sensitive to surface roughness. On butt-welded joints (t = 10 mm) made of S960QL, an improvement in fatigue strength was observed by blast-cleaning in the low cycle and finite life region, although there is also a notch effect due to the weld transition, see Hensel et al. (2019). 3.5.4. Findings on components with holes subjected to normal stress The test series on the flat bars with central hole made of S235JR in the as-rolled condition (#17 and #19) and with additional blast-cleaning treatment (#18 and #20) show the lowest test-related scatter during the fatigue tests at R = 0.1, which, in combination with the investigations from section 3.3, confirms a reproducible process of blast cleaning over the entirety of the samples, although this is not a standardized procedure or manufacturing process. Furthermore, the S-N curve predicted based on the FKM approach reliably describes the test data, although with some conservatism. DC 90 (m 1 = 5) appears to be too conservative for the selected component geometry. Furthermore, for series #19 the test data for different mean stresses (varying R-ratios) provide comparable numbers of cycles N f regarding the fracture events at R = 0.1, which shows the mean stress independence for this steel grade. For notched components, the FKM-Guideline (2020) specifies a surface treatment factor for shot peening of K V = 1.10 … 1.50. However, it also shows that blast-cleaning leads to a reduction in fatigue strength. At comparable stress levels, the fatigue life is reduced by around half due to blast-cleaning. This results in improvement factors k Imp < 1. The authors attribute the reduction to insufficient work hardening at the hole edge with a combined increase in surface roughness, which is in line with findings from Hertel (1969). A different conclusion was reached by Bannister et al. (2006), in which shot peening perpendicular to the hole wall with a special device had a considerable positive influence on fatigue strength. For a better understanding, metallographic microsections were taken from the specimen of series #18 to assess the work-hardened microstructure caused by the blast-cleaned area of the drilled hole. Fig. 6 shows the specimen preparation, which is divided into two different areas. In section A the edge of the hole and in section B the