PSI - Issue 75

Jörg Baumgartner et al. / Procedia Structural Integrity 75 (2025) 120–128

121

2

Jo¨rg Baumgartner / Structural Integrity Procedia 00 (2025) 000–000

1. Introduction

1.1. Influence of weld quality on the fatigue strength

The weld quality has a high influence on the fatigue strength of welded joints [4]. The reasons can be traced back to twoe ff ects: First, a notch acts as stress raiser and leads to a stress concentration. Main parameters for a fillet or butt weld are notch radius and opening angle, but also the position of notch (e.g. due to undercut) and the loading type, i.e. if normal or shear stresses act. Second, secondary bending moments need to be mentioned that can be caused by angular or axial misalignment of the welded joint. These misalignment increase the local stress amplitude. The quality of the weld seam is assessed through various geometrical parameters. Key factors influencing the quality of a fillet weld include the weld toe radius r and the weld toe angle α , which are defined according to ISO 5817 [5]. A small opening angle of α> 90 ◦ assigns a weld into quality level D, α> 110 ◦ is assigned to level C and α> 150 ◦ to level B, Table 1. It is important to note that this assignment does not necessarily correlate to fatigue properties. For a better correlation between weld quality and fatigue strength, Annex B of ISO 5817 can be used. In this table, a correlation between quality and fatigue strength is given, Table 2. As example, a fillet weld with an angle in the range 110 ◦ ≥ α > 150 ◦ is assigned a FAT90, i.e., an endurable nominal stress of ∆ σ n = 90 MPa. Few other quality related dimensions are also included in Annex B, such as the weld toe radius, Table 2. The values in this table have partially been adopted from a proposal by Hobbacher and Kassner [11]. In some standards, such as in EN 13001 [3] or EN 1999-1-3 [6], weld quality and fatigue strength are directly correlated, i.e. higher FAT-classes can be used if predefined quality levels are fulfilled. In other standards, a minimum required quality level B is directly coupled to the fatigue strength, as it is done, e.g., in Eurocode 3 [4]. FAT-classes cannot be increased (or decreased) with increasing (or decreasing) quality levels. An example of the strong influence of weld quality on fatigue strength is shown in Fig. 1. 3-point-bending load is applied on a transverse sti ff ener. A 3-point bending load is applied to a transverse sti ff ener, causing the highest stresses at the transition between the horizontal plate and the weld toe – also the location where the crack initiates. Although the “high quality” weld has a toe angle of α = 135 ◦ , it exhibits lower fatigue strength than the weld with excessive asymmetry. In this case, the reduction of stress concentration due to the lower toe angle in the highly stressed region between the horizontal plate and the weld is beneficial. Conversely, a smaller weld toe angle, as found in the weld with an incorrect toe, leads to reduced fatigue strength. The di ff erence in fatigue strength between specimens with excessive asymmetry (beneficial in this applied loading mode) and those with an incorrect toe amounts to about 30 % at N = 10 6 cycles and even 50 % at N = 5 × 10 5 cycles, due to the di ff erence in knee point and slope. This underlines the importance of considering weld quality in fatigue strength assessments.

1.2. Measurement of weld geometry

For the quantification of weld seam quality, assessing local geometrical parameters is crucial. The classical ap proach, primarily using manual weld gauges, is described in ISO 17637 [2].Besides this common solution, the evalu ation of geometrical parameters via 3D scans has gained increasing importance in recent years. Modern 3D scanners allow the e ff ective digitization of weld surfaces with high accuracy, while the evaluation based on such highly accu rate scans o ff ers a user-independent assessment of weld quality [18]. Currently, two major sensor types are available

Reference to ISO6520-1

Imperfection designation

Limits for imperfections for quality levels t (mm) D C B

No.

Remarks

α ≥ 90 ◦

α ≥ 110 ◦

α ≥ 150 ◦

1.12 505

Incorrect

weld

≥ 3

toe

Table 1: Definition of incorrect weld toe listed in Table 1 of ISO 5817 [5]