Meso-scale energy harvester: a comparison between MEMS and micro-machined designs

With smaller, cheaper, and more energy-efficient electrical components, energy harvesting systems have been a more attractive source of energy supply for wireless sensors, transducers, and other devices. One great example of mostly unused energy is the vibration of industrial machines. Along with the rise of predictive maintenance, more wireless sensors have been used to monitor those machines. Where the vibration energy present in those machines can be used to extend the sensor's life constrained by the battery. This work presents two fabrication approaches to design these devices using the piezoelectric principle: MEMS fabrication and micro-machined devices. MEMS are widely investigated for harvesting purposes for their capability of building complex micro-scale structures (< 0.1 cm(3)). However, it can be difficult to designing MEMS energy harvesting systems for the low frequency range (40 Hz to 200 Hz), which is the operating range for standard industrial machines. The adapted micro-machined harvesters from off-the-shelf piezoelectric components mostly used in macro-scale applications (> 10 cm(3)), can be an alternative in this situation. Numerical models were developed to simulate the dynamic behavior of the piezoelectric device and used as input for design optimization. The models were improved using a differential evolution algorithm optimizing in terms of the Normalized Power Density (NPD) and Mechanical stress. In order to validate these models, prototypes were built ns tested, with the results compared considering the NPD and frequency bandwidth. The optimization process raised key design aspects of meso-scale low-frequency piezoelectric devices, including stress limits of thin-film piezoelectric and fabrication complexity, Overall, these aspects suggest that there is an advantage of micro-machined designs over MEMS devices for these applications.