ANALYSIS OF SLENDER PIEZOELECTRIC WING CONFIGURATIONS FOR ENERGY HARVESTING: AEROELASTIC MODELING AND EXPERIMENTAL COMPARISONS

The study of the aeroelastic behavior of slender piezoelectric wings gains its relevance within the design perimeter of High Altitude Long Endurance (HALE) aircraft, or more in general of energy independent systems. As a matter of fact, the exploitation of new energy sources, without implying any direct penalization of the flight performances nor of the original aircraft design concept, is particularly appealing for HALE unmanned air vehicles (UAVs). Long range missions entail several design requirements such as high aspect ratio wing and low zero fuel weight, both with the common objective of reducing the energy consumption. However, albeit the structural design challenges afforded during the last years to increase as much as possible the mission duration of HALE aircraft, satellite systems still remain the most effective solution for ground surveillance purposes. Therefore, having additional energy from alternative sources, such as from structural vibrations, has to be embraced as a mission evolution opportunity for HALE UAV. The research activity presented in this thesis aims at investigating the energy extraction, via the application of piezoelectric patches over the wings' surface, from the most commons aeroelastic phenomena: critical flutter, sub-critical and super-critical LCOs, and gust response. For the sake of the just mentioned study, a nonlinear analytical and numerical aeroelastic piezoelectric wing model, which includes geometrical nonlinearities up to the third order and a multi-modal approach, was developed. The importance of higher order nonlinear terms is furthermore investigated via a comparison with FEM and experimental results. The numerical results, in agreement with the experimental results, shown that when the wing undergoes to high static deformations a state of dynamical instabilities may settle at speed even 50% lower than the critical flutter speed, and, in particular, when the oscillation amplitude becomes high, the model has to include higher order nonlinear terms to correctly capture the real oscillation amplitude. The results in terms of energy harvesting from the gust induced vibrations shown that the Squared gust seemed to be more effective for energy harvesting purposes than the 1-Cosine, if compared on the base of the energy content subtended by each curve of the gust profiles. Furthermore, although the 1-Cosine appeared less effective in terms of the amount of power that it can provide to the wing for energy harvesting, it was identified an optimal value of gust penetration gradient at which the assumed piezoelectric wing was able to extract the maximum amount of energy. Finally, thanks to the modal shaker and wind tunnel tests campaign, the importance of the location of the piezoelectric patches over the wing with respect to its dynamical response was investigated. What was seen is that the amount of extractable energy, at LCO, from the second bending mode of the wing is higher than that extractable from the first bending mode and it increases if the piezo-patches are slightly moved towards the wing center. This results suggest the necessity to develop a piezoelectric wing with multiple piezoelectric patches properly located in order to extract energy from the higher number of modes, or simply to the most excited mode, according to the good knowledge of the operational wing dynamic behavior. The order of magnitude of the maximum instantaneous power extracted from the assumed model during LCOs is of 10 mW, a good result if compared to the power demand of modern electronic devices.