Vibration energy harvesting is an attractive technique for potential powering of wireless sensors and low power devices. While the technique can be employed to harvest energy from vibrations and vibrating structures, a general requirement independent of the energy transfer mechanism is that the vibration energy harvesting device operate in resonance at the excitation frequency. Most energy harvesting devices developed to date are single resonance frequency based, and while recent efforts have been made to broaden the frequency range of energy harvesting devices, what is lacking is a robust tunable energy harvesting technique. In this paper, the design and testing of a resonance frequency tunable energy harvesting device using a magnetic force technique is presented. This technique enabled resonance tuning to ±20% of the untuned resonant frequency. In particular, this magnetic-based approach enables either an increase or decrease in the tuned resonant frequency. A piezoelectric cantilever beam with a natural frequency of 26 Hz is used as the energy harvesting cantilever, which is successfully tuned over a frequency range of 22-32 Hz to enable a continuous power output 240-280 μW over the entire frequency range tested. A theoretical model using variable damping is presented, whose results agree closely with the experimental results. The magnetic force applied for resonance frequency tuning and its effect on damping and load resistance have been experimentally determined.

Vibration energy harvesting is being pursued as a means to power wireless sensors and ultra-low power autonomous devices. From a design standpoint, matching the electrical damping induced by the energy harvesting mechanism to the mechanical damping in the system is necessary for maximum efficiency. In this work two independent energy harvesting techniques are coupled to provide higher electrical damping within the system. Here the coupled energy harvesting device consists of a primary piezoelectric energy harvesting device to which an electromagnetic component is added to better match the total electrical damping to the mechanical damping in the system. The first coupled device has a resonance frequency of 21.6 Hz and generates a peak power output of ∼332 μW, compared to 257 and 244 μW obtained from the optimized, stand-alone piezoelectric and electromagnetic energy harvesting devices, respectively, resulting in a 30% increase in power output. A theoretical model has been developed which closely agrees with the experimental results. A second coupled device, which utilizes the d 33 piezoelectric mode, shows a 65% increase in power output in comparison to the corresponding stand-alone, single harvesting mode devices. This work illustrates the design considerations and limitations that one must consider to enhance device performance through the coupling of multiple harvesting mechanisms within a single energy harvesting device.

Future deployment of wireless sensor networks will ultimately require a self-sustainable local power source for each sensor, and vibration energy harvesting is a promising approach for such applications. A requirement for efficient vibration energy harvesting is to match the device and source frequencies. While techniques to tune the resonance frequency of an energy harvesting device have recently been described, in many applications optimization of such systems will require the energy harvesting device to be able to autonomously tune its resonance frequency. In this work a vibration energy harvesting device with autonomous resonance frequency tunability utilizing a magnetic stiffness technique is presented. Here a piezoelectric cantilever beam array is employed with magnets attached to the free ends of cantilever beams to enable magnetic force resonance frequency tuning. The device is successfully tuned from −27% to +22% of its untuned resonance frequency while outputting a peak power of approximately 1 mW. Since the magnetic force tuning technique is semi-active, energy is only consumed during the tuning process. The developed prototype consumed maximum energies of 3.3 and 3.9 J to tune to the farthest source frequencies with respect to the untuned resonance frequency of the device. The time necessary for this prototype device to harvest the energy expended during its most energy-intensive (largest resonant frequency adjustment) tuning operation is 88 min in a low amplitude 0.1g vibration environment, which could be further optimized using higher efficiency piezoelectric materials and system components.

No abstract