PSI - Issue 79

Martin Sladký et al. / Procedia Structural Integrity 79 (2026) 421–432

426

Table 1. Characteristic parameters of the thin-walled fillet-welded specimen configurations comprising the experimental dataset.

Total number of specimens

Accepted number of specimens

Wall thickness t [mm]

Load asymmetry ratio R

Assigned nominal FAT nom

Assigned hot-spot FAT hs

Configuration

Source

CHS lap CHS fillet

12 14 13

10 11 10

1.24 1.24

0.1 0.1 0.1 0.1

69 78 78 78 78 78 88 88 78 78

193 Machacˇ et al. (2022) 193 Machacˇ et al. (2022) 193 Jiao et al. (2013) 193 Jiao et al. (2013)

CHS-Plate fillet S

1.6 1.8

CHS-Plate fillet G 6

6 3 2 5 9 7 8

RHS fillet OPB RHS fillet IPB Plate fillet T Ben Plate fillet T Ten Plate-RHS fillet

3 2 6 8 8

2 2 2 2 2 2

0 0 0 0 0

193 Gurney (1997) 193 Gurney (1997) 110 Gurney (1997) 110 Gurney (1997) 110 Gurney (1997) 99 Gurney (1997)

0.1

13

Plate fillet L

Twenty-node hexahedral elements were used throughout the finite element models, with a small number of other solid quadratic elements employed only where geometric constraints required. Dedicated submodels were imple mented following the recommendations of Cormier et al. (1999) to satisfy the requirements for hot-spot and notch stress evaluation specified by Hobbacher and Baumgartner (2024). Hot-spot stresses were determined as type “a” hot-spot stresses, as defined by Hobbacher and Baumgartner (2024), using a fine mesh and linear surface stress extrapolation from points located at thickness-relative distances of 0 . 4 t and t from the weld toe. According to Hobbacher and Baumgartner (2024), the recommended maximum element edge length at the hot spot is 0 . 4 t , with any additional refinement left to the analyst’s discretion. In this study, the finite element meshes featured a characteristic element edge length of approximately 0 . 1 t , corresponding to absolute dimensions ranging from 0 . 124mm to 0 . 20 mm. Figure 3 shows a representative mesh used for the hot-spot stress evaluation and schematically illustrates the linear surface stress extrapolation procedure.

σ

t

σ hs

0.4 t

0.1 t

Fig. 3. Representative finite element mesh used for hot-spot stress evaluation, illustrating the stress extrapolation from surface nodes.

For the notch stress evaluation, Hobbacher and Baumgartner (2024) specified that at least three quadratic elements should be placed over a 45 ◦ arc at the notch root to control the numerical error, an issue investigated in detail by Baumgartner and Bruder (2013). Given the reference notch radius of 0 . 05 mm adopted for all configurations, this entails a maximum finite element edge length of approximately 0 . 012 mm. In this study, each submodel dedicated to notch stress evaluation employed a graded mesh, with element edge lengths gradually decreasing from about 0 . 02mm at the submodel boundaries to 0 . 006 mm at the notch root, as illustrated in Figure 4, which provides a detailed view of a representative notch stress mesh.