This study aims to examine the shear performance of fused deposition modelling (FDM)-printed polylactic acid (PLA) infill topologies and evaluate whether percentile-based region-of-interest (ROI) rankings from finite-element analysis are consistent with ASTM D5379 Iosipescu shear tests.
Eight slicer-native infill patterns with comparable material volume were analyzed in Abaqus using 10 × 10 mm unit-square models extracted from the ASTM D5379/D7078 shear zone. For each topology, ROI-based p95 values of shear stress, shear strain, von Mises stress, displacement magnitude and apparent shear modulus were calculated for open-edge and perimeter-shell conditions. Two perimeter-shell configurations, rectilinear 0° and 45° grid, were then manufactured as full-scale PLA ASTM D5379 specimens, tested under quasi-static loading and simulated numerically for comparison.
Adding a perimeter shell changed the load path by allowing the infill and wall to act together as a load-carrying frame, reducing stress concentrations and compliance. The 45° grid with perimeter shell gave the most favorable overall response, whereas triangular patterns performed worst because of acute-junction stress concentrations. Experiments confirmed the numerical ranking: the 45° grid specimens showed higher initial stiffness and peak load than the rectilinear 0° specimens, and the finite-element model reproduced the initial linear response with good agreement.
The study presents a percentile-based ROI framework that converts full-field finite-element results into stable design metrics and shows that rankings from reduced unit-square models remain consistent with full-scale simulations and experiments.
