PSI - Issue 82

Peter Haefele et al. / Procedia Structural Integrity 82 (2026) 174–181 Peter Haefele and Patrick Schwarz / Structural Integrity Procedia 00 (2026) 000–000

178

5

probability of crack initiation at the rolled surface, see Figs. 2(a) and 2(b). This phenomenon can be explained, on the one hand, by deformation-induced roughening of the rolled surface (Schwarz and Haefele (2022, 2025)) and, on the other hand, by changes in the residual stress state at the laser-cut edge. Residual stress measurements on laser-cut electrical steel specimens conducted by Gottwalt (2023) indicate that, despite the absence of carbon and phase transformation in the material, high tensile residual stresses are present at the laser-cut edge. It is argued that the higher cooling rate during laser cutting affects the microstructural condition of the edge, leading to grain refinement. The reduction of tensile residual stresses, combined with the convergence of average surface roughness due to plastic deformation, results in a comparable damage potential for the laser-cut edge and the rolled surface. In contrast, strain controlled tests show a different behaviour. Crack initiation at the rolled surface is observed only for specimens with strain amplitudes ε a ≤ 0.2 %. It is assumed that the limited number of tests, together with the similar damage susceptibility of both surfaces under plastic deformation, causes this statistical artifact. A distinctive feature of laser-cut specimens made from NO30-15 is the high occurrence of intergranular crack initiation, see Fig. 2(a), amounting to approximately 50 % (compared to about 25 % for polished edges and 0 % for shear-cut edges). No load dependence is observed. The occurrence of intergranular crack initiation is not consistently reported in the literature. While Gottwalt (2023) observed exclusively transgranular crack initiation for shear-cut edges, Schayes et al. (2016) identified a load-dependent behaviour, whereas the studies of Dehmani et al. (2016) and Du et al. (2017) did not reveal any clear correlation. Schayes et al. (2016) further noted that the probability of intergranular crack initiation increases with the silicon content of the material, although the general validity of this observation remains to be verified. Regarding the mechanisms leading to intergranular cracking, Gottwalt (2023) and Schayes et al. (2016) describe comparable phenomena. On the one hand, a pronounced anisotropy between neighbouring grains results in high incompatibility stresses at the grain boundaries. On the other hand, strain distribution within the grains is heterogeneous, exhibiting local maxima near the grain boundaries, which leads to dislocation pile-up at these boundaries. Observations on NO30-15 laser-cut specimens show that the fracture surfaces of intergranular cracks are predominantly oriented orthogonally to the loading direction in the tests. This is interpreted as an influence of grain orientation on the likelihood of intergranular crack initiation.

Fig. 2. Laser edges NO30-15 R σ = 0.1; (a) intercrystalline crack in the edge; (b) transcrystalline crack in the laser edge.

The load dependence of crack initiation sites for shear-cut specimens made from NO30-15 can be observed under both stress-controlled and strain-controlled conditions, see Figs. 3(a) and 3(b). For tests with σ an ≤ 190 MPa (corresponding to 96 % of the yield strength) under Rσ = 0.1, crack initiation occurs predominantly at the smooth-cut portion of the edge (Schwarz and Haefele, 2025). Investigations by Schwarz and Haefele (2022) on NO30-15 indicate that this behaviour can be explained by an inhomogeneous hardening of the shear-cut edge. The smooth-cut region exhibits a lower degree of hardening compared to the fracture zone. Consequently, the fatigue strength of the fracture zone is higher, leading to failures at the smooth-cut region under sub-yield stress levels. Under plastic deformation, however, the hardness of both regions becomes more homogeneous, resulting in random failure locations (Schwarz and Haefele, 2025). In strain controlled tests with R ε = – 1, crack initiation occurs in the smooth-cut region for ε a ≤ 0.4 % in about 60 % of the specimens, which is consistent with the observations from stress-controlled tests. At higher strain amplitudes, however, failures occur exclusively in the fracture zone of the edge. This can be attributed to the greater cold deformation of the fracture region during the shearing process, which results in higher strain hardening. At low numbers of cycles, the fatigue life of the specimen is dominated by the residual deformability of the edge (Dittmann and Paetzold, 2018). The increased hardening of the fracture zone can thus be interpreted as an indicator of reduced