PSI - Issue 57

5

Matthias Winkler et al. / Procedia Structural Integrity 57 (2024) 510–517 Matthias Winkler / Structural Integrity Procedia 00 (2019) 000 – 000

514

HFMI could be performed on all 3 details. However, in the case of the transverse stiffener, the web plate hindered the process in the cut-out area, see Figure 6.

Figure 6. HFMI-treatment on transverse stiffener: a) complete peening trace around the outer side of the stiffener, b) incomplete peening trace around the cut-out because of poor accessibility through the web plate

2.3. Results of reference tests To determine the number of load cycles at D = 1.0 (N D=1.0,95%,aw ) at the end of the design working life, the characteristic fatigue strength Δσ C i.e. the tolerable stress range at 2 million load cycles (95% survival probability, prediction bound) of the 3 investigated structural details had to be defined. Therefore reference tests have been carried out in the as-welded state without post-weld treatment. In Figure 7 are given the S-N-curves of the 3 details in as welded condition.

c

b

a

Figure 7. S-N-curves of investigated structural details – a) rail clamp, b) transverse stiffener, c) lateral attachment – in as-welded condition

2.4. Results of fatigue tests and processing speed measurements In Table 2 are given the results of the performed fatigue tests. N D=1.0,95%,aw is the number of load cycles to pre damage the specimens as simulation of the end of the design working life. D treat shows the relation between the total load cycles after the post-weld treatment and N D=1.0,95%,aw . During the treatments the processing speed was recorded to allow an economic evaluation. The mean values of D treat and the processing speed are summarized in Table 3. With an average extension of service life by a factor of 28.3 burr grinding showed the greatest effect. However, the processing speed is the lowest at 3.48 cm/min. HFMI led to an average extension of service life by a factor of 16.1 with the highest processing speed at 22.5 cm/min. No cracks initiated in the area of the cut-out (incomplete peening trace) of the transverse stiffener. The S-N-curves are given in Figure 8.