A multi-stable ultra-low frequency energy harvester using a nonlinear pendulum and piezoelectric transduction for self-powered sensing

This paper presents the design, theoretical modelling and experimental validation of a quad-stable energy harvester for harnessing ultra-low frequency random motions using a nonlinear pendulum and piezoelectric transduction. The multi-stable pendulum is created by the magnetic forces between magnets on the pendulum and a tip magnet on a piezoelectric cantilever beam. Two attractive and one repulsive magnetic forces in combination with the gravitational force of the pendulum create multiple stable positions for the pendulum. The multi-stable dynamics allow the pendulum to effectively convert low-frequency random kinetic motions from the host, e.g. human motion or wind turbine tower oscillation into the pendulum oscillation, enabling effective plucking of the piezoelectric beam with enhanced output power. A theoretical model, including the magnetic interaction, piezoelectric conversion and pendulum dynamics, is established to describe the electromechanical dynamics of the whole harvester. A prototype is fabricated and tested on a linear shaker at ultra-low frequencies (1–3 Hz) to showcase the capability of the harvester and to validate the theoretical results. Around 8 μW Root-Mean-Square output power was obtained at 2.5 Hz and 0.8 g of excitation. Using the experimentally validated theoretical model, a parametric study was carried out to examine the influence of different structural and operating parameters, such as pendulum mass and length, magnetic coupling strength, excitation frequency and amplitude, on the output power and operating frequency range of the energy harvester. The operation frequency range and output power can be effectively adjusted by changing the above-mentioned parameters. The self-powered sensing capability is then illustrated by integrating the harvester with an off-the-shelf power management circuit and a 22 μF storage capacitor. The capacitor was charged from 2.8 V to 4 V in 90 s, showing its capability of implementing battery-free wireless sensing for different Internet of Things (IoT) applications.