PSI - Issue 77

Roman Hofmann et al. / Procedia Structural Integrity 77 (2026) 237–247 Roman Hofmann et al. / Structural Integrity Procedia 00 (2026) 000–000

244

8

The scan strategies demonstrate that simple resorting or segmentation (thereby resulting in alterations to the thermal history) can achieve a reduction in residual stress. In contrast, the Voronoi strategy showed an unfavorable increase in distortion when a contour scan was applied around each Voronoi cell. When this contour exposure was omitted, however, the resulting deflection was comparable to the reference. These findings highlight that the benefits of Voronoi partitioning depend strongly on the specific implementation, particularly on whether additional contour scans are used. The indirect influence of the contour scan suggests the necessity of optimizing the overlap or the gap between the Voronoi cells.

Reference Linear Hatch Pilger Lenght 2mm

1 . 2

Reorder Index-3 Reorder Index-2 Voronoi without Poly-Hatch Voronoi with Poly-Hatch Time Reorder

1 . 0

0 . 8

0 . 6

z (mm)

0 . 4

0 . 2

0 . 0

0

10

20

30

40

50

x (mm)

1 Fig. 7. Cantilever beam deflection of 316L cantilever manufactured with di ff erent scan strategies - conventional linear scan, pilger, index reorder (with index 2 and 3), time reorder and voronoi scan (with and without hatch of each voronoi cell).

In addition to residual stress analysis, the influence of scan strategies on fatigue performance was investigated. The fatigue specimens were produced from the alloy Specialis ® on an SLM ® 280 system. The applied scan strategies include: Index Reorder with index 2, Time Reorder with a threshold of 10 µ s, and the Voronoi strategy. Conventional linear hatching served as reference. The number of cycles to failure is summarized in Figure 8. The reference specimens fabricated with the linear strategy exhibited a large scatter in fatigue life, indicating significant variability under otherwise identical processing conditions. In contrast, the specimens built with the novel scan strategies showed markedly reduced scatter, with the results for Index Reorder and Time Reorder nearly overlapping due to the similarity of their reordering sequence. The Voronoi strategy yielded comparable, and in tendency slightly improved, fatigue performance relative to the reorder methods. Although only a limited number of Voronoi fields could be applied within the specimen cross-section, the strategy appears to have achieved a comparable homogenization of the thermal history. Overall, the novel scan strategies not only reduced the variability observed in the reference specimens but also demonstrated the potential to improve reproducibility in fatigue-critical applications. A more comprehensive evalua tion, however, would require an increased number of test specimens and a detailed assessment of surface roughness in the gauge section, as this factor is known to strongly a ff ect fatigue performance.

4. Conclusion and Outlook

This work presented the development of new scan strategies, enabled and deployed by a custom Python-based slicer that provides full vector control. Four scan strategies: Index Reorder, Time Reorder, Voronoi, and Pilger are presented with the goal of improving the thermal homogeneity and change the mechanical properties and structural integrity of PBF-LB / M manufactured parts.

• Process insight via thermography: Infrared snapshots of a demonstrator geometry revealed that conventional linear hatching produces pronounced heat accumulation toward the end of a scan path (especially when the