A multi-method investigation of high-aspect-ratio natural fiber-SMA composite wings: flutter delay, vibration suppression, and nonlinear surrogate modeling

High-aspect-ratio natural-fiber composite wings are attractive for lightweight and sustainable aerospace applications, but their relatively low stiffness and limited intrinsic damping make them susceptible to aeroelastic instability, large vibration amplitudes, and bounded nonlinear oscillations under aerodynamic loading. To address these limitations without imposing a significant mass penalty, this study investigates the aeroelastic response of banana-fiber composite beams with and without embedded superelastic NiTi shape memory alloy (SMA) reinforcement through a combined linear flutter-prediction, experimental, and reduced-order modeling framework. Computational flutter analysis was carried out in MSC Nastran SOL 145 using unsteady aerodynamic loading and the P–K method to evaluate modal damping evolution and flutter-related trends. These numerical predictions were complemented by wind-tunnel experiments on cantilever-mounted specimens, where vibration response was measured using accelerometer-based data acquisition and interpreted through maximum-displacement trends, frequency-response characteristics, and phase portraits. Three configurations were examined: NF R40, NF R60, and NF SMA. The results showed that passive material modification significantly influences the aeroelastic response within the investigated velocity range. Within 7–10 m s⁻¹, NF R60 exhibited the most consistent reduction in displacement amplitude and the most compact bounded phase trajectories, indicating stronger suppression through stiffness–mass modification. In contrast, NF SMA produced only a small mass increase relative to NF R40 while still providing clear response attenuation at lower velocities, demonstrating a more weight-efficient response modification. To obtain interpretable reduced-order descriptions of the measured dynamics, a forced Duffing surrogate and a physics-informed Van der Pol–Duffing model were identified directly from the experimental data. The Duffing model captured the softening nonlinear response effectively, while the PINN/VINO framework reproduced the dominant periodic behavior and bounded limit-cycle structure. Overall, the study establishes an integrated workflow for evaluating hybrid bio-smart composites and demonstrates the potential of SMA-reinforced banana-fiber laminates for passive aeroelastic response tailoring.