PSI - Issue 75

Kevin Mouradian et al. / Procedia Structural Integrity 75 (2025) 616–624 Kevin Mouradian et al. / Structural Integrity Procedia (2025)

621

6

3.4. Quality levels of specimens An overall assessment of the quality levels associated with each type of imperfection is presented, along with a final quality level for each specimen. The reported values for weld toe angle and weld toe radius represent the mean values across both sides of the weld. For undercuts, the mean value of the larger undercut measured between the two sides of the weld is considered. Table 4 presents the quality level of each imperfection detected within the weld. In addition, each specimen is assigned two overall “Final QL”; one based on the general ISO 5817 (CEN 2023a) tolerance limits (regardless of fatigue considerations), and another based on the stricter tolerances outlined in Annex B for fatigue- loaded welds. The “Final QL” represents the lowest quality level found within the imperfections. Assessment against the fatigue limits revealed that all specimens fell below quality level C63, including the BW_TYPE specimens (except BW_TYPE_2), which were intended to verify quality level B. These findings highlight the notable sensitivity of welded connections to imperfections. All specimens exhibited a significant angular misalignment between the plates, with values ranging from 1.3° to 3.1°, corresponding to quality level C63 and lower. The angular misalignment was oriented such that both plates inclined toward the weld surface. As a result, during clamping in the fatigue test setup, a pretension was introduced along the weld surface.

Table 4. Overview of weld imperfections and final quality level

Linear mis. h ali [mm]

Angular mis. α ali

Excess weld h w [mm]

Weld radius r [mm]

Weld angle α 172.2 170.3 173.2 163.2 165.8 164.9 166.1 174.9 165.3 167.2 166.0 168.0 [°]

Porosity d l

Final QL annex B

LOF l [mm]

Undercut h und

Final QL ISO 5817

Specimen

[°] 2.3 1.4 2.4 3.0 2.1 3.1 2.1 1.3 1.9 1.6 1.8 1.0

[mm]

[mm]

TYPE_1 TYPE_2 TYPE_3 TYPE_4 TYPE_5 TYPE_6 UND_1 UND_2 UND_3 UND_4 UND_5 UND_6

0.2 0.1 0.2 0.0 0.2 0.0 0.2 0.2 0.1 0.5 0.6 0.5

0.76 1.61 0.62 2.09 1.55 1.74 0.58 0.43 0.58 0.65 1.50 1.15

10.6 10.1 10.2

1.1 6.0 5.0

- -

0.52 0.11 0.32 0.01 0.03 0.02 1.48 0.54 1.21 1.20 0.85 0.80

D B D D

C63

9.4 5.8

3.3 5.6 6.7 3.2 4.0 4.0 4.2 2.7 2.5

11.0 40.0

- - - - -

1.3 1.2

B

26.0

6.2

15.0 60.0

8.8

- -

-

D

Imperfections that verify QL B and also B90 Imperfections that verify QL B but C63 Imperfections that verify QL C and also C63 Imperfections that verify QL D and < C63 Imperfections that verify QL < D and < C63

Imperfections that verify B90 Imperfections that verify C63 Imperfections that verify

4. Fatigue test results The fatigue tests were carried out in the laboratory of the Civil Engineering department of École Normale Supérieure Paris-Saclay (ENS PS). The experiments employed a MTS fatigue testing machine, with a maximum load capacity of 250 kN. The specimens were clamped at both ends with hydraulic wedge grips. An axial tensile load was applied with a stress ratio of R = 0.1 for all tests. The tests were performed at a loading frequency of 10 Hz. Only tests with a number of cycles to failure within the range 10 4 ≤ N ≤ 5×10 6 were considered for statistical analysis. Table 5