PSI - Issue 75

Xiru Wang et al. / Procedia Structural Integrity 75 (2025) 85–93 Wang / Structural Integrity Procedia 00 (2025) 000 – 000

86 2

1. Introduction The weld toe is usually the fatigue-critical location of each full penetration weld (Hobbacher and Baumgartner, 2024). An extensive number of theoretical and experimental studies have confirmed that the fatigue strength of the welded joint is directly tied to the local weld geometry at the weld toe (Ning Nguyen and Wahab, 1995; Lieurade, Huther and Lefebvre, 2008; Barsoum and Jonsson, 2011; Schork et al. , 2017, 2020; Hultgren and Barsoum, 2020). This relationship emerges due to the stress concentration, or notch effect, at the weld toe. The stress concentration factor (SCF) measures this notch effect and is influenced by the local surface geometry and load case (Zerbst et al. , 2018). This necessitates a swift and reliable SCF quantification for weld quality control. Usually, the weld toe radius and weld toe angle and plate thickness are used to calculate the SCF at a weld toe, as those parameters have been shown to be most influential (Kiyak, Madia and Zerbst, 2016; Schork et al. , 2017, 2020). Various SCF solutions are available for different types of welded joints. Comparisons between SCF evaluated using analytical solutions and SCF evaluated directly using the finite element (FE) method by Ottersböck et al. (Ottersböck, Leitner and Stoschka, 2021) and Schubnell et al. (Schubnell et al. , 2020) show much higher variation in the results of the FE calculations than the analytical solutions predicted, especially in the case that features, such as undercuts, are included that have not been considered in the derivation of the relationships. According to the current version of ISO 5817:2023 Annex B, the irregularity of ‘incorrect local weld toe’ is defined by weld toe radius and weld toe angle , see Figure 1 (a), in three quality levels: C63, B90 and B125. The quality levels are coupled with the corresponding fatigue classes FAT63 and FAT90 (Hobbacher and Baumgartner, 2024). A weld toe radius from < 1 mm corresponds to FAT63 while > 1 mm corresponds to FAT90, according to this definition. Hobbacher and Kassner (2012) propose different quality levels for the irregularity of the ‘incorrect local weld toe’ for fillet welds based on the weld toe angle of FAT63 (D), FAT71 (C) and FAT80 (B), with the quality level in bracket. Furthermore, there is currently no unified definition regarding how the parameter and are supposed to be determined. Investigations show the potential of 3D-scans for quality assessment of welded joints (Schubnell et al. , 2020; Renken et al. , 2021). However, round robin studies on the determination of and show high differences between the participants (Renken et al. , 2025). The differences may be caused by non-unified procedure for the determination of the geometrical parameters. Previous investigations related to this work by Wang et al. , (2025) show a weak correlation between local weld quality and fatigue life at transverse stiffeners with start-stop positions, as discussed section 2. For the reasons, the quality assessment of welded joints based on the local geometry presents several challenges. Thus, the aim of this work is to investigate the influence of local weld quality and stress concentrations on the fatigue strength of the joint. For this, a reverse engineering approach based on the translation of 3D-scans to Finite Element Models is used, to direct evaluate the stress concentrations based on the local weld geometry. For the evaluation of 3-D FEA result, the effect of different a/c ratios during crack propagation, and statistical size effects should be considered. 2. Quality Assessment and fatigue tests In this work, single-layer transverse stiffeners, manufactured by Gas Metal Arc Welding (GMAW) were used. The base material was S355. Details about manufacturing process and base material are given by (Wang et al. , 2025). The specimen was stress-relieved at 560 °C for 30 minutes to assure that residual stresses were excluded from further investigation. The specimens were welded with several start-stop-positions (intermittent welds) to create a variation of the local geometry, shown in Figure 1 (b). The specimens were digitalized with a stripe light projection (SLP) sensor SC3500-120 with a point spacing of 35 µm. The parameters weld toe angle and weld toe radius were evaluated based on 2D cross sections, extracted in a distance of 1 mm from 3D-Scans, see section 4, illustrated in Figure 1 (b). A smaller slice distance does not change