Perpustakaan Digital ITB

Porous lattice structures are widely studied as a way to reduce stiffness mismatch in loadbearing implants. However, preparing lattice models for finite element analysis can still require repeated manual steps, especially during geometry generation, mesh preparation, boundarycondition setup, and result extraction. This study aims to develop a voxel-native modelling workflow that can generate lattice structures from user-defined parameters and prepare them for finite element analysis. It also aims to verify whether the resulting voxel-based models can produce elastic responses that are consistent with previous studies. The developed workflow consists of a Python-based voxel lattice generator and an automated Abaqus compression-analysis script. The generator creates a three-dimensional voxel grid, evaluates the selected TPMS field, controls the target porosity, removes small disconnected artefacts, and exports the solid voxels as hexahedral finite element meshes. The Abaqus script imports the generated input file, assigns material properties, creates the compression setup, applies boundary conditions, and extracts force-displacement data to calculate the effective Young’s modulus. The verification case focused on a Strut-Based Gyroid structure at 30% relative density with lattice orientations of [100], [110], and [111]. The workflow produced effective Young’s modulus values of 109 MPa, 145 MPa, and 181 MPa for the [100], [110], and [111] orientations, respectively. The results reproduced the expected anisotropic stiffness order and remained within a reasonable range compared with numerical and experimental reference values. The observed deviations were mainly attributed to differences in geometry-generation convention, orientation alignment, and loading implementation.