PSI - Issue 75

Georg Veile et al. / Procedia Structural Integrity 75 (2025) 184–192 Georg Veile / Structural Integrity Procedia (2025)

190

7

In (Veile et al. 2025) the FDP RM and FDP FFS showed non- conservative predicted fatigue life with a Δy of 0.05 mm. In this study, the influence of gradient assessments by means of Δy on the predicted fatigue life using gradient based FDP was investigated. Increasing the Δy to 0.1 and 0.2 mm led to more realistic results in this work, especially with FDP hat showed non-conservative results in previous work. With increased Δy the non -conservative runout decreased to a predicted fatigue life within a factor of 10. This observation coincided with the other data points in the LCF and HCF regime . With increased Δy the predict ed fatigue life increased to the conservative green area. The superior FDP FGF showed less sensitivity to changing gradient values. Nevertheless, the results became also more conservative. However, the scatter remained higher compared to the scatter of fatigue life prediction with idealized weld radius of 1 mm. A different approach in the assessment of gradients does not reduce the scatter but non-conservative results. 4. Conclusion and outlook While different approaches to determine gradients are the solution to stabilize FDP FFS and FDP RM , the challenge to reduce scatter in the prediction of fatigue lifetime of scanned weld topologies could not be solved. Furthermore, smoothening the scanned weld topologies from radii of up to 0.2 mm to 0.55 mm did not affect the precision or scatter of the advanced, gradient-based FDP. However, common FDP, such as FDP FS or FDP SWT , showed a reduction of scatter with increased radii. Hence why, it is evident that the increase in scatter is based on geometrical factors, that are not considered by multiaxial consideration of shear- and strain-gradients. Nevertheless, the findings are in accordance to reference (Jung et al. 2025), where smaller radii lead to less accuracy than radii with increased magnitude. In an additional study, the results of radii smoothened up to 1 mm and above will be investigated. Furthermore, other factors, describing the true local weld radius will be investigated on their capability to reduce scatter in the fatigue life assessment of scanned weld topologies. Acknowledgements As corresponding author, I would like to thank my student Daniel Klöss, as well as my former student and now colleague Julius Lotz, for helping me with investigation of fatigue life assessment of scanned weld topologies. Furthermore, we would like to thank Dänekas et al. for sharing their algorithm to determine the weld toe radius. The authors thank the Federal Ministry for Environment, Nature Conservation, Nuclear Safety and Consumer Protection (BMUV), Germany for the financial support of this research project VHCF-II (1501636A and C). References Braun, Moritz; Ahola, Antti; Milaković, Aleksandar - Saša; Ehlers, Sören (2022): Comparison of local fatigue assessment methods for high-quality butt-welded joints made of high-strength steel. In: Forces in Mechanics, 6, p. 100056. DOI: 10.1016/j.finmec.2021.100056. Dänekas, Christian (2024): cdaenekas/WeldAnalysis: Initial Release. DOI: 10.5281/ZENODO.12805727. Dänekas, Christian; Schubnell, Jan; Krautheimer, Johannes; Jung, Matthias; Ghafoori, Elyas; Schaumann, Peter (2025): Algorithms for determination of weld toe radius and weld toe angle in welded joints. In: Journal of Constructional Steel Research, 224, p. 109112. DOI: 10.1016/j.jcsr.2024.109112.