PSI - Issue 77

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

243

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

7

uniform surface condition over most of its height. However, it must be noted that the first 8 mm of this sample were built using a conventional linear scan strategy. The transition between the two approaches is clearly visible as a change in surface coloration, underlining the distinct thermal influence of the applied strategy. This observation, which was made incidentally, suggests that the optimization of process parameters on reference geometries, such as cubes or standard tensile specimens, cannot be directly transferred to more complex components. In the cantilevers, the rib regions represent a reduction in cross-sectional volume, which modifies the accumulated heat input and results in locally di ff erent surface. The influence of scan strategy on the thermal homogeneity within a single layer is illustrated in Figure 6. The same geometry, an impeller, was produced using (a) a conventional linear scan strategy and (b) the Voronoi scan strategy, enabling a direct comparison. The thermal images were captured immediately after the exposure of the part which was manufactured by one subsequent contour scan after the inner region. Consequently, the bright outline and visible sparks in the images correspond to the contour scan rather than the bulk hatching. It should be noted that these snapshots do not represent the maximum temperatures at each point reached during processing; rather, they highlight the accumulation of residual heat within the component. For the linear scan strategy, heat clearly accumulates toward the end of the scan path. In particular, the right-hand side of the impeller shows a pronounced temperature gradient to the left-hand side and a temperature accumulation, indicating the risk of localized overheating. In contrast, the Voronoi strategy produces a more distributed temperature field. The three last zones of completed Voronoi fields can be identified, with no significant heat build-up is observed in any single region. As a result, the overall component is exposed in a more homogeneous manner, avoiding the steep temperature gradients seen in the linear case. Therefore, the presence of small temperature gradients within each of the Voronoi cells is evident, thereby allowing each cell to be identified within the thermal camera’s image.

a b Fig. 6. In-situ thermography of impellers manufactured with di ff erent scan strategies: (a) conventional linear, (b) Voronoi. Snapshot taken immedi ately after exposure. Exposure of part ended with a single contour scan. The thermographic observations provided qualitative evidence of heat accumulation and cooling behavior, which can influence microstructure and mechanical properties. To further investigate these e ff ects, cantilever specimens were fabricated using di ff erent scan strategies. Similar comparisons of parameter and conventional scan strategies on residual stress have been reported in previous work [11]. In this study, the cantilevers were produced from 316L stainless steel on an Evobeam SLaVAM 300 system. A conventional linear scan strategy was used as reference. The developed strategies were implemented as follows: Index Reorder with indices 2 and 3, Time Reorder with a threshold of 10 µ s, Pilger with a minimum segment length of 2 mm, and Voronoi both with and without contour exposure of each cell. The resulting cantilever deflections are summarized in Figure 7. The reference specimen showed a maximum upward deflection of 1.186 mm. The time-based reordering resulted in a slight increase in distortion, although not statistically significant. This e ff ect can be attributed to the relatively simple cantilever geometry, in which the time based algorithm produced only limited to no reordering in the upper layers of the part (above the internal support structures). The index-based reordering showed a moderate reduction of residual stresses, with deflections of 1.03 mm (Index 2) and 1.06 mm (Index 3). The Pilger strategy exhibited the largest e ff ect on reducing deflection to 0.899 mm.