PSI - Issue 57

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

456

5

In future, longer longitudinal attachments can be categorized better according to the upcoming Eurocode (prEN 1993-1-9, 2023). It should be noted, however, that the weld must be welded in a single pass around the ends of the attachments. If the weld is ending or beginning at the already critical attachment end, a significantly lower fatigue strength due to a sharper notch may be the result. Yet, if the weld is going all around the edge, the modified classification in the upcoming Eurocode is just as safe and allows for a more economical design of rectangular longitudinal attachments without post-weld treatment. On the other hand, the fatigue strength can be enhanced considerably, irrespective of the length, if the ends of the attachments fulfillthree conditions: if they have a minimum radius of 150 mm,if they are fully welded through in the area of the radius, so there is no longer a critical weld root, and if the welds transition is ground flush. All three requirements are quite cost-intensive, the radius due to the high material offcuts, the through -welding and the grinding due to the high workload. The fact that grinding the weld transition zone has a positive influence on fatigue strength is well known and can also be seen in many other details, such as the transverse butt joint. The better fatigue strength of a detail without a weld root in areas with high stress concentration is also known. Until now, though, there have been no tests to show whether the radius of the longitudinal attachment has a positive effect on the fatigue strength. These tests now were carried out as part of the project (Feldmann, et al., exp. 2024). To determine not only the influence of the attachment radius but also to enable the classification of this detail into a detail category, 18 tests were carried out. Except for the radius at the end of the attachments, all tests were realized under the same conditions. The length of the attachment was 500 mm. The ends were welded around in one go and fully penetrated, afterwards the weld transition zone was grounded off. Nine specimens, each with a radius of r = 80 mm and 150 mm, as seen in Fig. 4 a) and b) were tested. The specimens were loaded with different stress ranges whereby the crack initiation was observed either at the transition to the attachment or the base plate, irrespective of the radius. 3.2. Influence of the radius

1000

100 Stress range Δσ [MPa]

10

100.000

1.000.000

10.000.000

Cycles [N]

r = 80 r = 150

Detail Category 80 MPa Detail Category 100 MPa

b)

a)

c) Fig. 4: Longitudinal attachments with a radius of a) r = 80 mm, b) r = 150 mm c) Experimental results (Feldmann, et al., exp. 2024). A very low scatter of all results, as seen in Fig. 4 c), is reflected in a strong correlation of the test results R Pearsons = 0.88.A joint evaluation of the test results is therefore reasonable.The experimentalvalues adapt well to the slope of m = 3 and reach Δσ 95% = 100 MPa when evaluated both individually and together. The crack initiation of the only exception with fewer cycles was caused by the weld root of a weld that was not thoroughly welded. This was therefore not considered in the evaluation.