RESUMO
Laser powder bed fusion (LPBF) provides a rapid and versatile approach for producing parts with complex geometries. However, many parts with intricate geometries have overhang structures, which are not easily fabricated by using LPBF and are often downgraded by staircase effects, warpage, cracks, and dross formation. Thus, the present study proposes a combined numerical and experimental approach for determining the optimal settings of the laser power and scanning speed that minimize the surface roughness and maximize the density of Inconel 718 LPBF overhang structures. In the proposed approach, the heat transfer simulations are employed to determine the melt pool depth, the melt pool length, and the solid cooling rate within the feasible input space of laser power and scanning speed combinations. Notably, the simulations take account of both the difference in the material properties of the solid and powder materials, respectively, and the variation of the laser absorptivity in the depth direction of the powder layer. The simulation results are then used to train artificial neural networks for predicting the melt pool depth for 3600 combinations of the laser power and scanning speed within the input space. The resulting processing maps are screened in accordance with three quality criteria (namely the melt pool depth, the melt pool length, and the solid cooling rate) to determine the optimal processing region, which improves the surface roughness. The feasibility of the proposed approach is demonstrated by fabricating 10 × 10 and 20 × 20 mm2 horizontal overhang structures using parameter settings chosen from the optimal processing map. It shows that the optimal processing conditions result in a low surface roughness and a maximum density of 99.78%.
