A Polygon-Based Method to Minimize Layer Underfill Voids by Optimizing Raster Angle in Material Extrusion Additive Manufacturing

Abstract

Achieving 100% infill density in material extrusion (MEX) additive manufacturing is crucial for producing high-performance components. However, this technology is inherently prone to generating voids that compromise mechanical properties. Specifically, the common use of raster (rectilinear) infill and contours often produces underfill voids - such as corner, edge and gap voids - whose geometries are highly dependent on filling parameters, particularly the raster angle. Although void formation is a significant challenge, research has primarily focused on inter-bead porosity as a function of process parameters, leaving the systematic optimization of toolpath parameters to minimize layer underfill voids relatively unexplored. To address this gap, this work introduces a novel parallel polygon-based method to identify the raster angle that minimizes layer filling voids. The proposal analyzes raster orientations over a discretized angular domain (0°–179° in 1° increments) and also considers interlayer interleaving through a 90° rotational offset. To cope with the high computational cost associated with fine angular discretization and complex geometries, the main part of the algorithm is implemented using GPU-based parallel computing. Tested with various geometries, the proposed method successfully quantifies projected void areas and identifies the optimal raster orientation within the discrete search space. The results confirm that the projected void area is significantly affected by part geometry and filling parameters, specifically bead width, raster gap, and, most notably, raster angle. The developed method serves as an effective approach to reducing toolpath-induced voids through the optimal selection of the raster angle.