Multi-objective optimization and energy absorption mechanisms of bio-inspired lightweight sandwich structures based on Pareto representative solutions

Abstract

To develop localized lightweight load-bearing and energy-absorbing components for UAV fuselages and wings, this study proposes a bio-inspired multilayer Ti-6Al-4 V sandwich structure based on beetle elytra. A parametric unit cell with an elliptical cavity and hollow cylindrical struts is established, using core wall thickness, face-sheet thickness and core height as design variables. A Box–Behnken RSM–NSGA-II framework is used to balance nominal density and energy absorption, yielding three Pareto-representative designs: Min-ρ, Knee and Max-EA. Specimens with one to three layers are fabricated by selective laser melting and tested under quasi-static compression. Prediction errors for nominal density and energy absorption are within 6%. Total energy absorption increases with layer count, but specific energy absorption (SEA) depends strongly on design type and multilayer stacking. Max-EA reaches the highest SEA of 50.90 kJ·kg⁻¹ at L = 2 but decreases to 44.57 kJ·kg⁻¹ at L = 3 because of diagonal shear fracture and rapid densification. In contrast, Knee-L3 shows a buckling–shear coupled collapse mode and achieves 49.26 kJ·kg⁻¹, surpassing Max-EA-L3 while maintaining a flatter plateau and lower load fluctuation. Finite element comparisons of the three Pareto-representative L3 configurations show that the model can reasonably capture the main load–displacement responses across different collapse modes, with the Knee-L3 deformation sequence further validated by experimental images and PEEQ contours. These results show that the Knee design provides a representative compromise between lightweighting, energy absorption and multilayer stability, offering guidance for additively manufactured energy-absorbing structures.