Geometric fidelity dominates dynamic prediction accuracy in Bio-Inspired FWMAV wings: a validated 3D-DIC and FE modelling study

Purpose This paper aims to establish a noninvasive, high-fidelity workflow for capturing 3D geometry and structural dynamics of bio-inspired flapping wings, enabling accurate dynamic prediction for flapping-wing micro air vehicles (FWMAVs). Design/methodology/approach An integrated approach combining projected-speckle 3D digital image correlation (3D-DIC) with image-based finite element (FE) modeling is developed to generate digital twins of ultra-thin wings. Dragonfly hindwings serve as representative structures to compare conventional fluorescent sprayed speckles with a novel computer-generated projected-speckle method in terms of geometric accuracy, measurement noise, and mass loading. Findings The projected-speckle technique reduces 3D point-cloud noise by 37%, improves corrugation accuracy to within 3% of ground truth and adds no measurable mass. FE models from projected-speckle geometries predict the fundamental frequency within 2.1% of experiments (157 Hz versus 153.8 Hz) and achieve Modal Assurance Criterion values above 0.87 for bending and coupled modes, whereas sprayed-speckle-based models overestimate frequency and fail to capture torsional coupling. Inverse identification returns vein and membrane elastic moduli of 6.1 ± 1.2 GPa and 1.5 ± 0.3 GPa, respectively, showing that geometric fidelity dominates predictive accuracy. Originality/value This work introduces and validates a non-contact projected-speckle 3D-DIC and image-based FE pipeline that enables simulation-driven design of high-performance FWMAV wings, moving beyond empirical trial-and-error.