PSI - Issue 57

Gloria Hofmann et al. / Procedia Structural Integrity 57 (2024) 452–460 Hofmann, G.; Bartsch, H.; Kuhlmann, U.; Feldmann, M.

458

7

The fatigue loading of the cruciform joints was carried out with a stress ratio of R = 0.1 under constant tensile stress amplitudes of different stress levels. Complete fracture was defined as the failure criterion. It was found that the size of the residual gap and the size of the weld affect the failure location of the cruciform joint: The smaller the internal imperfections and the larger the weld, the more likely the weld toe is the decisive failure location. With large gap sizes and small welds, on the other hand, the weld root becomes decisive (Bartsch & Feldmann, 2021), (Bartsch & Feldmann, 2022). The tests could be successfully recalculated numerically applying the effective notch stress concept acc. to (Hobbacher, 2016). Based on validated models, a study can be carried out to determine the decisive failure location depending on the local weld geometry. Fig. 5 shows a limit curve showing the transition from root failure to toe failure as a function of the local weld geometry, for weld angles of 45° and plate thicknesses of 10 mm. According to this, geometries that lie above the curve exhibit toe failure and do not have to be designed for root failure. The imperfection can be tolerated in these cases. 4.3. Investigation of transverse stiffeners with incorrect root gap for fillet welds The influence of a remaining gap between components, also referred to as incorrect root gap for fillet welds, was investigated by fatigue tests on transverse stiffeners welded in beams (non-load carrying joint) with weld imperfections. Three series of transverse stiffeners with different gap sizes were performed for this purpose. The stiffeners with a thickness of 10 mm were welded with 5 mm fillet welds into the rolled HEA240 beams with a steel grade of S355 and in case of large gaps of 3 mm, a welded beam with same dimensions. During fabrication, it has been noticed, that some of the large gaps were partly filled by filler metal of the welds. The fatigue loading of the transverse stiffeners was carried out with a stress ratio of 0.1 under constant bending stress amplitudes of different stress levels. Failure initiated in all cases from the weld toe. Complete fracture of the tension flange was defined as the failure criterion. As can be seen from Fig. 6, even large gaps did not lead to a reduction of the fatigue strength of the transverse stiffener. The scatter of the test results is slightly higher for the series with a ga p size of 3 mm, which included a welded beam instead of a rolled beam, as for the other two series. All test data show clearly higher DC than currently prescribed in (EN 1993-1-9, 2005) and (prEN 1993-1-9, 2023). Even large gaps of 3 mm thickness did not lead to a significant reduction of fatigue strength.

ͶͲ ͸Ͳ ͺͲ ͳͲͲ ͳʹͲ ͳͶͲ ͳ͸Ͳ

Δσ transformed to 2 mio. cycles [MPa] gap thickness s [mm] –‡•– ”‡•—Ž–•

’ ” ͳ ͻ ͳͻͻ͵ͻǦ ͳ͵ ǦǦ ͻͳ ǦȀͻ

Ͳ ͳ ʹ ͵ Ͷ

Fig. 6: Fatigue strength (average value for 50% probability of failure) of transverse stiffener depending on gap thickness.

4.4. Summary

As the results of the investigations described above show, not every weld imperfection must have a critical effect on the fatigue behaviourof a member. For this reason,further investigations were carried out into the influence of weld imperfections. As a result of H. Bartsch's dissertation,the Fatigue Class Combination Model(FCCM) was developed,