Adaptive path planning for additive manufacturing via convex decomposition and partial-contour-guided hatching

Path planning is a critical challenge in additive manufacturing, as it directly affects part quality and process reliability. Conventional algorithms often struggle with complex geometries, leading to voids, overlaps, and discontinuities that compromise structural integrity. This paper presents an adaptive path-planning framework that combines convex decomposition with a partial-contour-guided hatching (PCGH) strategy to generate continuous, space-filling, and geometry-conforming toolpaths. The proposed approach segments complex interior regions into fillable sub-polygons using a collision-aware concavity analysis and tree-search procedure, while PCGH constructs efficient hatch paths by selectively leveraging boundary segments as guidance. This integration ensures geometric conformity, minimizes defects such as overlaps and voids, and maintains path continuity across intricate features. The framework is demonstrated and validated in wire arc additive manufacturing (WAAM), which is a directed energy deposition (DED) process where continuous deposition, collision avoidance, and stable layer formation are essential. Tests on diverse geometries, from simple polygons to curved and multi-featured structures, confirm state-of-the-art performance in both simulations and physical experiments. By addressing a long-standing limitation in path planning, this work enables more reliable and defect-resistant additive manufacturing of complex parts.