Aerodynamic testing of a smart composite wing using fiber-optic strain sensing and neural networks

• Published in: Smart materials and structures 2000-12, Vol.9 (6), p.767-773
• Main Authors: Lunia, Abha, Isaac, Kakkattukuzhy M, Chandrashekhara, K, Watkins, Steve E
• Format: Article
• Language: English
• Subjects: Acoustics; Exact sciences and technology; Fundamental areas of phenomenology (including applications); Noise: its effects and control; Physics
• ISSN: 0964-1726; 1361-665X
• DOI: 10.1088/0964-1726/9/6/305

The feasibility of developing a smart wing using composite materials, fiber-optic sensors, and neural networks is investigated. Strain and aerodynamic lift force induced in a wing model at different air speeds and angles-of-attack are experimentally determined in an open circuit wind tunnel. The smart wing model consisted of a glass/epoxy composite beam with an interferometric fiber-optic sensor mounted at the wing root. The strains were correlated to the set of air velocities and angles-of-attack using a feed forward back propagation neural network approach. The resulting neural network simulation could predict the experimental strain in real time with an average error of 3.17%. Finite-element analysis prediction of aerodynamic lift calculated using measured strain at the wing root generally agreed with thin airfoil theory, the differences being attributed to uncertainty in composite material properties.