Partially decoupling cell size and density as tailorable design parameters in mSLA-printed gyroid cores for sustainable sandwich panels
Year: 2026
Type: preprint
Abstract: Lightweight sandwich cores can be tailored to the load case, yet design rules linking lattice architecture to core and panel performance are lacking for vat photopolymerisation. Sheet-network gyroid lattices were printed by masked stereolithography (mSLA) from a partially biobased PLA-based resin across a full factorial of unit cell size (5–20 mm) and nominal density (5–20%). The printable window, density, wall thickness and compressive properties were characterised. Selected cores were integrated into sandwich panels with basalt fibre facesheets in a partially biobased epoxy matrix and tested in flexure. Printability was bounded by wall separation below ≈ 130 µm and by resin entrapment in 5 mm cells at high density. Cell size controlled wall thickness (132–271 µm at 5% density), whilst apparent density was independent of cell size at low density (p ≥ 0.14), so the two act as partially decoupled design levers. Compressive strength followed bending-dominated Gibson–Ashby scaling (n ≈ 1.5 in the defect-free range and ≈ 2 across the full window). A cell-size effect on strength was robust at 20% density on raw and ρ²-normalised bases, and emerging at 15%. It is attributed to coupled wall thickness and cell count rather than the architecture itself. Panel flexural rankings matched core rankings. The 20 mm/20% core yielded the stiffest panels whilst the 5 mm/10% core gave the most strain-tolerant, benignly failing panels. These results establish mSLA gyroid lattices as tailorable sandwich cores, provide indicative design rules for core selection, and show that partially biobased constituents can form structurally efficient sandwich panels.
Source: Research Square
Institution: University of Cape Town
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Funder: National Research Foundation
OpenAlex ID: W7225913332
