PSI - Issue 75

Martin Edgren et al. / Procedia Structural Integrity 75 (2025) 555–563 Martin Edgren et Al. / Structural Integrity Procedia (2025)

561

7

7

Figure 6 a) Weld geometry section next to the introduced notch, b) section of the area of the notch – the red lines represent overlays of sections of the region of interest (roi) of the weld geometry, and the black line represents the centre contour of each roi, c) scatter of the test specimens for the metric Weld Toe Radius. The introduced notch has similar geometric properties as some parts of the as-welded geometry, i.e. the welded geometry competes with the notch geometry in terms of sharp transition localizing the crack initiation. Figure 6 a) and b) illustrate how the notch geometry, possibly is competing with the as-welded weld geometry in terms of sharp transition between parent material and weld. Furthermore, the Winteria ® scan revealed that the scatter within the specimen and between different specimens was quite significant with regards to the metric Weld Toe Radius, see Figure 6 c) , a geometric property shown having an impact on the fatigue strength [24]. It is suggested to further study different notch techniques to ensure proper crack initiation at the predefined positions in the test specimen. This is especially important with non-mechanized welded specimen (mimicking on site welds) often having a greater variability in terms of weld quality and geometric properties over the length of the weld [25]. The monitoring of the change in stiffness has been performed directly logging the voltage measurement, i.e. not using a calibrated strain value as input. This approach is lacking in information of the actual strain data and only referencing drop in voltage values over the strain gauge acting as a resistor. The approach prohibits the verification of the FEA using the measured data directly. In this case the study is relying solely on the stiffness change in percent projected with aid of the FEA. Furthermore, the measurement was conducted with equipment offering partial online monitoring, i.e. the measurements are not evaluated in real-time. It is suggested that future studies are conducted with calibrated strain measurements and an equipment capable of fully on-line real time monitoring to ensure strain feedback to FEA and ensure high speed data processing for better resolution of the implemented stop criterion respectively. 7. Conclusion • Introducing a notch in a specimen can be challenging depending on the geometry of the specimen and the shape and position of the notch. A method to introduce a mild notch in a gusset plate specimen has been presented. The notch has been introduced at the vicinity of the weld toe and between the specimens they have a depth of 0.15-0.35 mm. • A method for strain range drop monitoring has been developed. Even though this study was primarily conducted through fatigue experiments, a finite element analysis was utilized to estimate the strain drop over the strain gauges following the stiffness change for a crack propagating from the initial notch to a specific crack depth. The results from the FEA were used to choose a common strain range drop stop criterion value of 5%. The strain range drop of 5% project crack depths estimations between 0.4 and 0.6 mm. The following conclusions can be made: