Nonlinear Dynamics of Nitinol-Enhanced Carbon-Fiber-Reinforced Polymer Beam-Rod in Subsonic Flow

This study investigates the nonlinear aeroelastic behavior of carbon-fiber-reinforced polymer (CFRP) laminates embedded with prestressed pseudoelastic nitinol shape memory alloy (SMA) wires under subsonic flow. High-aspect-ratio beam specimens with 4-layer and 10-layer layups were tested in a subsonic wind tunnel across velocities from 10 to 31 m/s. Experimental analyses, including frequency response functions (FRFs), phase portraits, and tip displacement, were supported by numerical flutter simulations via the P-K method in MSC NASTRAN ® . CFRP-SMA specimens exhibited superior dynamic performance compared to plain CFRP and aluminum-reinforced counterparts, showing frequency shifts up to 10 Hz and amplitude reductions of 20–40%. The 10-layer SMA specimen achieved peak resonance at 323 Hz, while flutter frequency in the 4-layer specimen dropped from 14.04 to 11.6 Hz. Heating the specimens to 45°C further enhanced damping, reducing amplitudes by an additional 25%. For predictive modeling, LSTM and GRU neural networks were trained on experimental data, achieving real-time displacement predictions with [Formula: see text], outperforming physics-informed-neural-network-based models. These findings highlight the effectiveness of SMA integration for passive flutter control and validate the use of deep learning in digital twin development for adaptive aerospace structures.