Vibration prediction of a wind turbine gearbox using Fast Fourier Transform algorithm and Bond Graph methodology

Abstract

Several wind farms are located in different countries of the world to generate electricity from the kinetic energy of wind. These farms are still installed in isolated cities far from the centers of use. Several research efforts have been made to bring these wind turbines closer to their users in order to minimize the cost of generating and transporting electricity. But because of their size and weight, wind turbines face a lot of problems, the two main ones are noise and mechanical vibrations. The vibration induced by the movement of blades and nacelle causes considerable effects on the wind turbines themselves, on the infrastructure and on the ground. Nevertheless, the multiplier remains an important source of noise and vibration due to the meshing forces between gears. In this work, a physical model of a wind turbine gearbox with the horizontal axis is established by means of Bond Graph formalism. This model is exploited to predict vibrations within the transmission chain subjected to a variable aerodynamic torque coming from the rotor using the FFT algorithm.