MODELING AND ANALYSIS OF A VIBRATION ENERGY HARVESTER UTILIZING FREQUENCY UP-CONVERSION FOR LOW AND VARIED SPEED OF ROTARY STRUCTURES

Abstract

This thesis developed a mathematical model for a two-degree-of-freedom energy harvester that converts low-speed mechanical rotation into a piezoelectric cantilevered beam vibration. The harvester utilizes the swing motion of a small disk mounted on a large structure that rotates at a low speed (e.g. wind turbine blade) to stimulate vibration a piezoelectric beam by a magnetic repelling force. A frequency up-conversion technique is used to transform the rotation frequency of the structure to the higher vibration frequency of the beam. The corresponding electromechanical model of the energy harvester is developed using the energy method by including magnetic repelling force and piezoelectricity as coupling terms. A system of three governing equations describes the motion of the disk, vibration of the beam, and voltage output of the harvester. These equations are solved using an ODE45 function in MATLAB software and the results are verified by the corresponding experimental study. The performance of the harvester is analyzed in two configurations: (i) the disk rotates in the rotation plane of the structure (in-plane) and (ii) the disk rotates normal to the rotation plane of the structure (normal-to-plane). The varied-energy-harvesting performance is studied at different rotational speeds. At low blade speeds, the harvester generates power through regularized magnetic excitation per blade revolution. Using the in-plane configuration, a more dynamic disk movement as well as a higher voltage and power are generated when the ratio of centrifugal acceleration to gravity is more than unity. At higher blade velocities, the increased centrifugal force ratio reduces the motion of the disk and the performance of the harvester decreases. In the normal-to-plane configuration, the effect of the centrifugal force is eliminated, and the swinging motion of the disk is driven only by the change of gravity. The results show that the model can predict the power peak as a function of blade speed, and the proposed harvester can generate a considerable amount of power for self-sustainable sensing and monitoring of wind turbine blades. Additionally, the effect and sources of the intermodulation distortion and harmonic distortion caused by nonlinearities of the mechanism in the voltage output of the harvester are described.