Low-Frequency and Broadband Vibration Energy Harvesting Using Base-Mounted Piezoelectric Transducers

Piezoelectric vibration energy harvesters often consist of a cantilevered beam composed of a support layer and one or two piezoelectric layers with a tip mass. While this configuration is advantageous for maximizing electromechanical coupling, the mechanical properties of the piezoelectric material can place limitations on harvester size and resonant frequency. Here, we present numerical and experimental results from a new type of piezoelectric energy harvester in which the mechanical properties and the resonant frequency of the cantilever beam resonator are effectively decoupled from the piezoelectric component. Referred to as a base-mounted piezoelectric (BMP) harvester in this paper, this new design features a piezoelectric transducer mounted beneath the base of the cantilevered beam resonator. The flexibility in the material choice for the cantilever beam resonator means that the resonant frequency and the beam dimensions are essentially free parameters. A prototype made with a 1.6 mm $\times4.9$ mm $\times20.0$ mm polyurethane beam, a PZT-5H piezoelectric transducer, and an 8.36-g tip mass is shown to produce an average power of 8.75 and $113~\mu \text{W}$ at 45 Hz across a 13.0- $\text{M}\Omega $ load under harmonic base excitations of constant peak acceleration at 0.25 and 1.0-g, respectively. We also show an increase in full-width half-maximum bandwidth approximately from 1.5 to 5.6 Hz using an array of four individual BMP harvesters of similar dimensions with peak power generation of $10.38~\mu \text{W}$ at 37.6 Hz across a 1.934- $\text{M}\Omega $ load at 0.25-g peak base excitation. Finite elements-based numerical simulations are shown to be in reasonable agreement with experimental results, indicating that the harvester behaves like a damped mass–spring system as proposed in this paper. Fabricated using casting and laser machining techniques, this harvester shows potential as a low-cost option for powering small, low-power wireless sensor nodes and other low-power devices.

I. Introduction

Thanks to the miniaturization of sensors and continued reductions in the power consumed by microcontrollers and wireless transmitters, it has become increasingly popular to place sensors in locations that are remote, difficult to access, or both [1]. Since it is often not possible to run wires to sensors in these locations, and access can even be so restricted that changing batteries is not feasible, researchers have been exploring ways to scavenge the energy required to power sensor systems from the sensors’ environment. Light, heat, and RF radiation are all common sources of energy that a system can scavenge energy from [2]. However, in some cases, such as monitoring the structural integrity of buildings and bridges [3], or the air pressure in a tire [4], the most abundant source of accessible energy is vibrations.