Numerical Simulation of the Bending Behavior of a Topologically Optimized Structure Printed in ABS by FDM: G-Code-Driven Mapping of Local Material Orientations

Abstract

This preprint presents an original dual-branch numerical pipeline for simulating the bending behavior of an MBB (Messerschmitt-Bölkow-Blohm) beam, topologically optimized using the SIMP method and manufactured in ABS by fused deposition modeling (FDM). Starting from a single G-code file generated in OrcaSlicer, two parallel processing branches are derived in Python: (i) a faithful 3D mesostructural reconstruction via the CadQuery library, producing STEP geometries for the architectural characterization of three studied raster orientations (0°, 45°, 90°); (ii) an automatic mapping of local deposition directions onto the centroids of the ANSYS mesh elements via a k-NN algorithm (KD-tree, O(N log N)), enabling the assignment of a locally oriented orthotropic elasticity tensor to each element through APDL commands (EMODIF, ESYS), using material constants from Somireddy and Czekanski (2017). The SIMP optimization produces a structure retaining 30% of the initial volume for a 6.6% reduction in bending deflection. The comparative simulation across orientations reveals that the 90° orientation offers the best mechanical performance (Dmax = 0.136 mm, -54.4% vs. 0°), consistent with the architectural mechanisms identified from the reconstructed mesostructures. This work has not undergone external peer review at the time of this deposit and is shared as a numerical baseline pending experimental validation. Python scripts and APDL command blocks are included in the appendices for reproducibility. This manuscript is currently under review at Rapid Prototyping Journal (Emerald).