Windbelt generators are an energy harvesting method that have viable uses for small-scale battery-charging applications for operation in low to medium wind speed conditions. The windbelt mechanism is a fluid-structure interaction (FSI) between airflow and a flexible belt, causing unstable vibration called aeroelastic flutter. The aim of this study is to understand the relationships between key parameters which will aid the future design of high efficiency windbelts. Experiments were conducted to test different length and chord rubber belts in a wind tunnel. It was found that the critical velocity of flutter onset and its frequency are a function of surface area and to a lesser extent belt pre-tension. In parallel, a novel linear FSI theoretical model that couples Theodorsen's method and Euler-Bernoulli beam theory was developed. The system is solved using a state-space method; thus we are assess the applicability of linear theory to predict the onset of non-linear limit-cycle flutter. The computational model captures the trends of the experimental data with varying quantitative accuracy and allows the identification of bending and torsional modes that couple to generate flutter. It is found that the model identifies either the bending or torsional mode that is unstable, the other coalescing mode is identified by examining the frequency space; it is the coupled bending mode that has the greatest influence on the flutter frequency. Better agreement for critical velocity is found at higher aspect ratios, demonstrating the over prediction of aerodynamic forces and absence of 3D effects at low aspect ratios in the Theodorsen method.

No abstract