PSI - Issue 75

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

560

6

detected — this stop criterion is derived from finite element analysis (FEA), corresponding to an estimated crack depth of 0.4 – 0.6 mm. This depth range accounts for variations in strain gauge placement and FEA predictions. Following the initial fatigue loading, specimens undergo High-Frequency Mechanical Impact (HFMI) treatment. After the post weld treatment, the depths of the grooves are assessed to ensure compliance with the International Institute of Welding (IIW) HFMI guidelines [5]. In the final stage, specimens are subjected to a second fatigue testing sequence, under a maximum load of 200kN, with a load ratio (R) of 0.1 and a frequency of 19 Hz, continuing until failure or until the run-out criterion is met. 5. Results The fatigue testing of the notched and prefatigued specimens are presented in Figure 5 showing fatigue life well above the characteristic nominal stress curve for HFMI improved welds (FAT90) for the notched, prefatigued and HFMI treated specimens. The data points for the current study are overlayed over 90% prefatigued results from [6] as a comparison. It is observed that only one of the specimens, specimen OB3, failed. Specimens OB2 and OB6 both reached 10,000,000 cycles, causing their results to coincide as a single data point in the graph below. It should be noted that run-out in a specimen corresponds to a double run-out, as notches are present on both sides of the specimen.

Figure 5. Results from fatigue testing. The notched pre fatigued specimens (green triangles) show fatigue life results far above the HFMI FAT 90 curve, and a reference specimen (red triangle) show fatigue life well above the FAT 50 curve. Datapoints from earlier study [6] are overlayed in current figure. 6. Discussion There are several techniques available to introduce a notch in a test specimen, e.g. wire EDM, hole drilling [22], [23]. However, for the specimen in this study, there are challenges with the mentioned approaches due to the geometry of the specimen. Firstly, the uneven weld geometry originating from the manual arc welding process and secondly difficulties reaching the weld toe due to geometrical restrictions, i.e. the shape and position of the gusset plate. As forementioned, this study developed and adopted a mild notch approach. When the geometry and/or the weld is well defined, a slight disturbance of the geometry could be sufficient [23]. However, the mild notch approach introduces specific challenges with respect to the introduced disturbance.