G-Code2STEP-Converter for accurate representation of geometries created by material extrusion

Purpose The purpose of this paper is to address the fidelity gap between original digital designs and manufactured parts in material extrusion additive manufacturing by developing and validating a software framework that generates accurate virtual models based on toolpath data. Design/methodology/approach A software framework, G-Code2STEP-Converter, was developed to generate accurate virtual models of material extrusion-produced parts based on toolpath motions. The framework’s performance was evaluated by conducting both physical and virtual three-point bending tests on specimens with varying infill patterns and densities. The virtual models were analyzed using finite element analysis. Findings Results from the physical and virtual three-point bending experiments exhibited strong correlation for all inspected infill patterns and densities, confirming that the new software framework provides high-fidelity digital representations of the manufactured parts. Research limitations/implications This study focuses on standard three-point bending tests and a range of commonly used infill patterns. Future work could extend the framework’s applicability to other testing methods and more complex geometries. Practical implications The proposed framework enables the creation of high-fidelity digital twins in additive manufacturing workflows, supporting process optimization, quality prediction and archiving of manufacturing intent. Originality/value This work introduces a novel tool for bridging the fidelity gap between designed and manufactured geometries in material extrusion, offering new capabilities for digital part verification and process documentation in additive manufacturing.