This study investigates the flexural performance and failure mechanisms of additively manufactured sandwich panels featuring 3D-printed corrugated cores and face sheets reinforced with E-glass fibers. The research aims to quantify the effects of core geometry, thermoplastic matrix selection and fiber reinforcement on peak load, stiffness and energy absorption, while critically evaluating the predictive accuracy of classical Gibson–Ashby models for extrusion-based composites.
Sandwich panels were fabricated using a modified dual-feed fused deposition modeling (FDM) process. To mitigate nozzle wear and facilitate the extrusion process, 100-tex continuous E-glass yarns were pre-coated with a styrene-butadiene rubber solution. Eight distinct sample configurations were produced and evaluated, utilizing two thermoplastic matrices (polylactic acid [PLA] and acrylonitrile butadiene styrene [ABS]) across three core cross-sections: rectangular, triangular and trapezoidal. These included both unreinforced polymer panels and continuous fiber-reinforced composites containing a ∼15% fiber volume fraction. Specimens were subjected to three-point bending tests in accordance with ASTM C393.
Experimental results demonstrated that structural performance is highly dictated by core geometry and matrix integrity. The trapezoidal composite panel (TRCP) achieved the highest peak flexural load (726.8 N) with stable post-peak progressive collapse. The rectangular composite panel (RCP) exhibited superior ductility and the highest specific energy absorption (SEA) at 118.9 N·mm/g due to progressive core wall buckling, dramatically outperforming the triangular core panel, which suffered a brittle, premature collapse at 558.8 N with an SEA of only 12.5 N·mm/g. Regarding matrix selection, PLA-based panels significantly outperformed their ABS counterparts (e.g. RCP peak load of 706.3 N versus RCA at 448.8 N) due to ABS’s susceptibility to severe intra-face layer delamination under compressive bending loads. Furthermore, the addition of continuous glass fibers improved the peak load by ∼19% and SEA by ∼24% compared to unreinforced PLA equivalents (RPP). Furthermore, while classical Gibson–Ashby theory predicted global core shear failure due to the high face-thickness-to-span ratio (t/L = 0.022), experimental observations revealed premature localized face indentation. Quantitative analysis confirmed that the printed cores only utilized 13% to 83% of their theoretical shear capacities, highlighting the critical limitations of applying continuous-foundation analytical models to discrete 3D-printed corrugated topologies.
This research provides robust quantitative benchmarking for continuous glass-fiber FDM sandwich structures. It demonstrates that classical sandwich theories require adjustment before application to 3D-printed composites, as they fail to account for real-world discrete support conditions, epoxy interface compliance and localized post-buckling behaviors. The findings offer vital, application-specific design guidelines for the geometric optimization of additively manufactured lightweight structures.
