Non-planar slicing for filled free-form geometries in robot-based FDM

Multi-axis techniques in Additive Manufacturing (AM) unlock promising features such as supportless fabrication, reduced material consumption, improved surface quality and mechanical properties. Among those techniques, non-planar (NP) slicing promises to be the most suitable approach to fabricate 3D-components with significant curvature such as free-form geometries. Those are characterized by a layer thickness variation (LTV) along the curvature, which should be minimized. Industrial 6-axis robots are mandatory to achieve such performances. NP slicing generalization is challenging. On one side, there is a need to define a suitable contouring method compatible with the different geometrical features present in objects. On the other side, the generalized slicing method must be able to reconstruct the inner side not provided by superficial information provided by triangular mesh. In this work, a new algorithm to generate filled NP layer has been proposed using a contouring method that reduces the LTV. The bidirectional rectilinear infill strategy has been adapted for NP layers providing a more feasible toolpath to Fused Deposition Modeling (FDM) and such Direct processes where curved paths are detrimental. The proposed strategy has been validated by fabricating a tubular geometry with a robotized FDM system. Tubular geometry provides a sub-optimal solution of LTV known analytically for the contour. The infill algorithm has been tested with a complex surface applying the NP torus on a waved shape. Previous studies consider only contour providing an LTV ranging in +0% ÷ -46%. This study considers only the inner side. The analytical LTV resulted in a range of +0% ÷ -60%. The cross sections of the components were analyzed and compared with the analytical results. Although the proposed infill strategy does not maintain completely the contour layer thickness in the infill side, it shows to be able to cover more complex NP layers without saddle points.