Investigation of Process-dependent Mechanical Behavior in Additive Manufacturing via G-code-driven Discrete Element Simulation

Abstract Finite element (FE)-based numerical approaches are widely used to simulate additive manufacturing processes and predict mechanical response, predefined meshes and repeated remeshing make it difficult to manage spatial resolution without adding complex computational overhead when simulating the path-following process. To address the challenges of mesh-dependent resolution, this study employs a bonded discrete element method (DEM) in which solids are represented as bonded particle assemblies whose bond network is defined by process parameters encoded in the toolpath and the discrete nature of the method allows for the direct representation of complex toolpath-induced internal architecture without frequent remeshing typically required in mesh-based approaches. Here, the additive manufacturing process is modeled as a sequence of G-code-driven particle deposition and bond-activation events, such that infill pattern, infill density, scan direction, and build orientation explicitly determine the resulting bond network. Firstly, we test the model’s capability in reflecting the effect of infill densities by using virtual three-point bending test by adjusting the bond’s parameters. Further, parametric studies are systematically conducted to investigate the coupled and individual influences of scan direction and build orientation across different infill configurations. It shows that the proposed method can reflect the changes in load-transfer paths, deformation localization, and sensitivity of bending stiffness to process parameters. These findings demonstrate that DEM-based virtual testing can link scan direction and build orientation to the elastic and flexural behavior of printed components, offering valuable guidance for developing additive manufacturing strategies.