PSI - Issue 75

Mattias Clarin et al. / Procedia Structural Integrity 75 (2025) 467–473 Clarin et al./ Structural Integrity Procedia 00 (2025) 000 – 000

470

4

2.3. Residual stress measurements 91 Residual stress measurements were performed in the vicinity of the weld toes using a Bruker D8 Discover 92 diffractometer with a 1 mm collimator. The measurement positions are shown in Fig. 3. 93 To determine the biaxial stress tensor, measurements were taken at three in-plane sample orientations: 0°, 45°, and 94 90°, as illustrated in Fig. 3. Due to interference from the protruding weld, it was not possible to perform measurements 95 in the opposite directions (180°, 225°, and 270°), which are normally included to account for shear stress components. 96 The measurement positions were located by monitoring the intensity of the 110 diffraction peak while translating 97 the sample stage from a clay-covered region of the weld, across the weld toe, and into the unwelded area. An increase 98 in peak intensity indicated that the measurement area had passed beyond the clay-covered region. Additional scans 99 between 40° and 50° 2θ, with a step size of 0.05 mm, were used to confirm that clay -related peaks were no longer 100 present in the diffraction spectra. 101 All measurements were carried out in iso- inclination mode using Cu Kα radiation, with a 2θ range of 134° to 140°. 102 The tilt angle Ψ was varied from 0° to 45° in ten discrete steps, corresponding to sin²(Ψ) values from 0 to 0.5. Each 103 measurement required approximately 14 hours. Residual stresses were calculated using the sin²(Ψ) method 104 implemented in the Bruker Leptos software. 105

106 Figure 3. Orientation and direction notations of the four residual stress measurement positions on the specimen in the as-welded 107 condition. Blue regions are covered with clay. 108 2.4. Fatigue testing 109 The experimental work was conducted using a high-frequency resonance testing machine, Rumul 500.All 110 specimens — excluding the overload cycling — were tested at a cycle frequency of approximately 80 Hz. To establish 111 a reference S-N curve for the welded detail without any applied overload (0% preload), a total of eight specimens 112 were tested to failure. 113 A frequency drop in a resonance testing machine indicates a loss of stiffness in the system. During fatigue testing, 114 this stiffness reduction is ideally associated solely with the test specimen and the growth of a fatigue crack large 115 enough to affect the dynamic response. The stop criterion (tripping frequency) was defined as a 3 Hz drop from the 116 initial test frequency for these tests. At this point, the crack was mostly visible to the naked eye, and the specimen was 117 deemed to have reached its fatigue life. 118 All tests were conducted under a load ratio of R=0.1. For the reference series (0% preload), nominal stress ranges 119 of 150, 200, and 300 MPa were applied. In the preloading series, the specimens were subjected to a single overload 120 cycle prior to fatigue testing, with overload levels corresponding to 25%, 50%, 75%, and 100% of the measured 0.2% 121 proof strength (981 MPa). The overload was applied in uniaxial tension using a servo-hydraulic testing machine, and 122 the stress levels were calculated based on the nominal cross-sectional area of the specimen. After the preloading step, 123 specimens were transferred to the resonance testing setup and tested at constant amplitude, with nominal stress ranges 124 between 200 and 400 MPa.

125 126