PSI - Issue 75

Mehdi Ghanadi et al. / Procedia Structural Integrity 75 (2025) 457–466 Mehdi Ghanadi et al./ Structural Integrity Procedia (2025)

458

2

1. Introduction Plate thickness in load-carrying welded structures subjected to fatigue loading can affect fatigue strength. Unlike the thickness effect, when the material thickness is less than the reference value, the welded joint experiences an increase in fatigue strength, a phenomenon referred to as the thinness effect (Gustafsson, 2006). Factors such as the size of the main load-carrying plate, transverse attachment, and weld size are key contributors. This influence, known as the size effect, can be interpreted from different perspectives. Statistically, larger volumes are more likely to contain defects. Since fatigue follows a weakest-link mechanism, more weak regions increase the probability of failure (Pedersen, 2019). The production process affects mechanical properties like residual stress and misalignment. In thicker plates, higher residual stress from thermal constraints, along with microstructure and surface roughness variations, represents the technological size effect (O. Öjasäter, 1995). Geometrical size effect expressed by stress gradient through thickness along crack propagation and stress concentration at the toe region. Decreasing plate thickness results in a steeper stress gradient through the thickness and reduced stress concentration, thereby increasing the fatigue capacity of thin plates (Pedersen, 2019). Based on the early work by Gurney (Gurney, 1991) and Maddox(Maddox, 1991; Schumacher et al., 2009), within the framework of theoretical and experimental investigations, the overall geometry of welded specimens has a significant influence on their fatigue strength, highlighting the size effect on fatigue performance. In their research, the thickness correction formula, including reference thickness and power exponent, was applied to account for the size-dependency of fatigue strength. Based on their study it has been shown that the size of the primary load-bearing plate, along with the transverse attachment and weld profile, is a key factor influencing the fatigue behaviour of welded joints. Mashiri et al. (Mashiri et al., 2007) studied the scaling effect concept, which means how parameters influence the fatigue strength, such as weld size, weld toe radius and plate thickness scale, concerning one another. Larsen et al. (Larsen, 2022) analysed the probabilistic and statistical framework of the thickness effect by investigating the stochastic size effect, considering the weld length and utilizing the stochastic stress concentration factor. In a further study, Ghanadi et al. (Ghanadi, Hultgren, Clarin, et al., 2024; Ghanadi, Hultgren, Narström, et al., 2024) found that the probabilistic model could decrease S-N scattering, indicating its relevance to thin plates. The study of uncertainty and scatter of fatigue strength, shown in the S-N curve, of welded joints arise from different sources, associated with material behaviour and residual stress, structural geometry, and local weld geometry (Fricke & Muller Schmerl, 1998). Stenberg et al. (Stenberg et al., 2019) show that fatigue behaviour varies by material despite similar geometry, and FAT curves alone are insufficient for thin, high-strength steel. By using fracture mechanics, Gustafsson (Gustafsson, 2006) found that welded joints with a plate thickness of 3mm are more sensitive to weld defects than thicker plates, which underscores the significance of weld quality measurement in thin structures. Welds are typically the weakest points in structures, with their quality directly influencing the overall integrity. In fact, weld quality is an important factor affecting fatigue life, while geometry and defects initiate crack formation (Barsoum & Jonsson, 2011). In thin plates, weld quality plays a crucial role in fatigue performance. High-quality welds help reduce stress concentrations by enlarging weld toe radii and angles, enhancing fatigue life (Ghanadi et al., 2025). The weld imperfections and defects cause stress to rise near the welds. Previous studies have made attempts to investigate weld quality by introducing the actual weld profile into the simulation. Hultgren et al. (Hultgren & Barsoum, 2020) investigated the correlation of weld geometry parameters, including leg length, toe radius, undercut and toe angle, with fatigue life by using laser-scanned weld geometry. In another study by Hultgren et al. (Hultgren et al., 2023), real weld geometry was incorporated in the fatigue assessment in combination with weakest link methods, resulting in reduction of the variation in fatigue life estimations. Weld fatigue life is improved by decreasing tensile residual stress or improving weld geometry. While post weld treatments are effective, they increase costs and lead time; however, optimising welding parameters, positioning, and material selection enhances fatigue strength without extra processing. In this context, Holmstrand et al.(Holmstrand et al., 2014) studied the geometry and quality of welds, considering toe weaving shape, showing increased fatigue strength with the weaving technique due to improved weld geometry as a result of the increased weld leg length and reduced flank angle which lower stress concentration at the weld toe, resulting in increased fatigue life. Åstrand et al. (Åstrand et al., 2016) showed that smooth undercuts in as-welded conditions can enhance fatigue strength analogous to post-weld treatments, decreasing production time while enabling high-quality, cost-effective welds. Post-weld treatments are categorized into weld profile modification and weld residual stress modification methods. Re-melting of weld toe by Tungsten Inert Gas (TIG) is a weld profile improvement technique that produces a smooth transition between the weldment and base material, minimizing stress concentration (Yildirim, 2015).