PSI - Issue 82

L. Mata et al. / Procedia Structural Integrity 82 (2026) 16–23

19

4

Mata et al. / Structural Integrity Procedia 00 (2026) 000–000

Fig. 3. Frist principal stress field at the notch region for the untreated material condition: (a) bending-torsion, σ⁄τ = 4⁄3; (b) bending-torsion, σ⁄τ = 2; (c) tension-torsion, σ⁄τ = 4⁄3; (d) tension-torsion, σ⁄τ = 2. White arrows represent the predicted initiation sites.

Regarding the generation of the bending-torsion moments, a single force (F) was applied to one side of the prismatic bar connected to one end of the specimen (see Fig. 2(a)). Both the normal and shear stress components were controlled by adjusting the ratio of d 1 to d 2 , keeping d 1 constant across all simulations. In the tension-torsion model, two opposing forces with equal magnitude (see Fig. 2(a)) were applied to generate the shear stress component, while the normal stress component was produced by applying a surface pressure applied to the concentric area of the gripping component connected to the end of the specimen. To validate the numerical approach developed in this study, experimental results reported by Cunha (2024) were used. Briefly, the tests were conducted under proportional bending-torsion loading using the specimen geometry displayed in Fig. 1 for both as-built and the stress-relieved material conditions. The notch was introduced after the additive manufactured process by CNC machining. Before testing, all the specimens were polished in the region around the notch with a set of sandpapers (P#500, P#1200 and P#2400) followed by the application of diamond paste. Two normal-to-shear stress ratios ( s / t = 4/3 and s / t = 2) were used, and the stress ratio (R) was set of 0. The nominal normal stress amplitude ( s a ) varied between 43.5 MPa and 23.1 MPa, while the nominal shear stress amplitude ( t a ) varied between 32.6 and 11.5 MPa. Fatigue life was defined as the number of cycles at which the crack length reached 3 mm.