PSI - Issue 76

Robin Motte et al. / Procedia Structural Integrity 76 (2026) 74–81

78

Fig. 3. S-N curves.

Figure 3 also shows S-N curves for S3555J2 (Ulewicz et al. (2014)) and bulk WAAMed EMK8 (Motte et al. (2024)) materials. For the WAAM material, specimens were extracted along the deposition direction (denoted as “X”) and along the build direction (denoted as “Z”). As these three reference curves were originally determined via rotating bending experiments with R = − 1, the Goodman mean stress correction was applied to convert that experimental data to R = 0 . 1 (Dowling et al. (2009)). The OLSLR parameters for these datasets are also listed in Table 3. The S-N curves for the CONV and CMT specimens show a clearly decreased fatigue life in comparison to the references (i.e. a lower intercept). Furthermore, the slopes are considerably steeper compared to the reference curves for polished specimens. The as-built surface waviness contributes to this behaviour as stress concentrations have a detrimental e ff ect on the fatigue life. Although the CONV specimens show a slightly steeper slope than the CMT specimens, these preliminary results do not suggest a considerable di ff erence in fatigue performance between the two remanufacturing processes. A point of attention is however that, according to ISO 12107, more fatigue tests should be conducted in order to obtain S-N curves for design or reliability purposes (International Organization for Standardization (2012)). These S-N curves can therefore only be considered as exploratory.

Table 2. Results of the fatigue experiments. Specimen ∆ σ nom [MPa]

Cycles to failure [-]

Cycles to crack initiation [-]

Fraction of total fatigue life until crack initiation [%]

CONV-1 CONV-2 CONV-3 CONV-4 CONV-5 CONV-6

150.98 219.60 200.39 180.07 160.31 260.23 200.39 219.60 260.23 180.07 160.31

Run-out 331 312 261 522 339 457 486 904 158 135 265280 242900 193631 496751 713565

-

-

168 515 50.86 52 697 20.15 120 049 35.37 101 008 20.75 55 195 34.90 87 684 33.05 86 468 35.60 12310 6.36 180 969 36.43 145 111 20.34

CMT-1 CMT-2 CMT-3 CMT-4 CMT-5

It was observed that, apart from a single run-out (no failure after 5 700 000 cycles) at a nominal stress range of 150.98 MPa, all test samples failed at the weld toe of the last deposited weld bead, regardless of the welding process. No imperfections such as pores or lack-of-fusion were observed on the fracture surfaces.