This paper presents the design of an electrostatic MEMS energy harvester for medical implant application. It examines solutions for constraining the motion of the mobile part in one direction and proposes an innovative spring architecture. Indeed, constraining the mobile part motion is essential to avoid undesired contact between comb electrodes. It is particularly important in environments in which mechanical vibrations result from complex combination of rotations and translations. The objective of the considered device is to power the next generation of leadless pacemakers using mechanical energy generated by heartbeat motion. Such solution would dramatically increase the lifetime of implants and would be very beneficial for the patients by reducing the number of replacement surgical operations. Numerical simulations based on analytical modelling and acceleration signal mimicking heartbeat motion enabled to analyze the system response in various condition, showing the interest and benefits of the proposed approach.

No abstract

Research on piezoelectric energy microgenerators from vibrations led to an abundant literature, with various strategies to optimize the frequency range and output power. In contrast, for very low frequency range (<10Hz) and/or for non harmonic mechanical source, the large majority of the strategies are not adapted. This work deals with a small scale piezoelectric generator where the input mechanical source consists of a single force application in the range of hundreds of Newtons (i.e., typical human weight). Contrary to vibrational mechanical sources, such an application context necessitates harvesting as much as energy as possible in a single cycle. This was achieved by assembling several piezoelectric stacks within a mechanical amplification system, and to use the electric field and stress levels close to the limits of the piezoelectric elements. Ericsson cycle (i.e. thermodynamic cycle comprising two iso-electric field and two iso-stress steps) was applied to the piezoelectric material and later using two device prototypes in order to quantify the harvesting capabilities. Finally, in a realistic application point of view, a passive electrical interface based on Bennet’s doubler was implemented and compared to the Ericsson cycles in terms of output energy. This electrical energy management strategy successfully allowed working at ultra high electrical field (>2kV/mm) enabling a converted energy density close to the ultimate value. An maximal energy density of 320 mJ/cm3 was reached using Ericsson cycles, and 130 mJ/cm3 using Bennet’s doubler (~40% of the ultimate energy density). The device comprising ~2.4 cm3 of piezoelectric material, the net output energy converted and stored per cycle reached 320 mJ. Still, the work presented here can be adapted to other range of forces and displacements for maximizing energy harvesting.