PSI - Issue 76

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

80

Fig. 5. Determining the crack length by: (a) considering the horizontal displacement field U obtained from DIC, (b) calculating the di ff erence in adjacent subsets, and (c) thresholding.

Fig. 6. Crack lengths as function of number of cycles.

was taken as the vertical distance from the bottom edge of the specimen. Before starting each fatigue test, a series of 10 images of the unloaded specimen was taken and subsequently analysed to obtain a metric for the noise in the displacement. From these static images, the di ff erence between the overall maximum and minimum values for U (i.e. the “worst-case” noise) was obtained and subsequently used as threshold value. The threshold values ranged between 0.8 µ mand2.0 µ m. The crack lengths for all specimens are shown in Figure 6. In Figure 5(c), it can be seen that the crack does not follow the 45° groove angle, but instead follows a shallower angle. An angle of approximately 10° was observed for all specimens. As the fatigue crack was seen to propagate in the S355 substrate material, the crack growth rate was governed by this material, rather than the WAAMed material. The fatigue performance of steel components remanufactured by WAAM with as-built surface waviness was in vestigated using four-point bending specimens specifically designed for this purpose. Specimens were manufactured using a conventional short-circuit welding process and a CMT process. Compared to smooth specimens of S355J2 and bulk WAAMed EMK8 material, the fatigue strength of the remanufactured specimens was seen to be negatively impacted by the surface waviness. Furthermore, the slopes of the S-N curves were seen to be considerably steeper compared to those for the polished reference material. Although the conventional process resulted in a somewhat steeper slope than the CMT process, the exploratory S-N curves for the two remanufacturing processes did not sug gest considerable di ff erences in fatigue performance. Based on the thermoelastic amplitude obtained through infrared thermography, it was determined that between 6 and 50% of the total fatigue life was in the crack initiation stage, 4. Conclusions