Skip to main content

Nonlinear Energy Harvesting Solutions
for Micro- and Nano-Technologies

Final Report Summary - NEHSTECH (Nonlinear Energy Harvesting Solutions<br/>for Micro- and Nano-Technologies)

Executive summary

Vibration energy harvesting at micro-scale is fundamental to enable low-power, maintenance free and long-lasting wireless sensor networks (WSNs). However, many problems still limit the efficiency of current vibration-driven generators: narrow bandwidth, low power density, MEMS scaling, inconsistent vibrating sources. The main scientific objective of this research project was to overcome present limitations by proposing an innovative dynamical approach. In particular, this idea is based on the exploitation of multiple-mass velocity amplification combined to nonlinear dynamical features such as bistable mechanical oscillator. A centimetre-scaled prototype of a electromagnetic harvester with buckled supporting beams was first implemented in order to investigate the dynamical behaviour and the harvesting efficiency when the system operates in the bistable regime. An enhancement in the bandwidth response of a factor of 2.5 along with an and higher generated power off its natural resonance was demonstrated when compared to the monostable regime. Based on this results, in the second phase of the project, the principle of a multiple mass oscillator which exploits the effect of mechanical frequency-up conversion was successfully validated by studying the behaviour of an electrostatic MEMS VEH mounted onto a bistable base suspended by buckled beams. Then, the final part of the project constituted the evolution of the proposed concept of applying the principle of multiple mass velocity amplification effect combined with bistability into a fully integrated MEMS VEH. A prototype of silicon MEMS VEH with electrostatic gap-closing combs was implemented. This prototype constitutes the apex of the NEHSTech project. Here, a free moving tungsten micro-ball was inserted inside the inertial silicon mass suspended by silicon serpentine springs. The nonlinear bistability of the silicon mass was controlled by the bias voltage as it operated a softening effect to the effective stiffness of the mechanical suspensions. In terms of power density and frequency bandwidth performances, this micro electrostatic VEH demonstrated an improving of more of a factor 4 to 10 in the power density with respect to existing linear micro generators. In addition, it shown to be capable of operating at very low frequency (10-60Hz) with a power density of around 20µW/cm3 that is comparable with that of a lithium battery operating for 1year. This technology could be suitable for peacemaker powering or other applications that can harness the kinetic energy of human movements. The training objective of the Fellow Researcher on MEMS modelling, design, manufacturing and characterization was very complementary to his existing skills. On the other hand, the researcher shared his know-how with the hosting group and disseminated the results in informal metting, coursed, workshops, international conferences and journal papers.

Summary description of project context and objectives

The harvesting of ambient energy in the shape of vibrations is considered an key technology to enable self-sustained, long-lasting and free-maintenance wireless electronics. Kinetic energy is abundantly available in industrial plants, transportations, infrastructures and also human activities. However, mechanical vibrations from natural and artificial sources are pretty much weak and located below few hundreds of hertz. On the other hand, typical vibration energy harvesters (VEHs) are resonant spring-mass-damper systems that only work efficiently at resonance frequency. In order to improve the bandwidth of harvesting devices some alternative concepts have been proposed in recent years: self-tuning resonators, piezoelectric cantilevers arrays and mechanical frequency-up conversion systems. Moreover, nonlinear Duffing-like piezoelectric oscillators have already been shown to be advantageous for harvesting energy under random noise and low−frequency vibrations. The conversion of mechanical energy at low frequencies has been also approached via frequency−up conversion techniques. However, piezoelectric and electromagnetic converters are still quite bulky, while electrostatic vibration harvesters are more suitable to be implemented into MEMS-scales. Nevertheless, at sub-centimeter dimensions the resonance quickly increases from few hundreds hertz to several kHz depending on the proof mass, thus deteriorating the capability to harvest energy at frequencies below 100 Hz. Recently, a multiple degrees of freedom (multi-DOF) electrostatic system using frequency-up conversion have been proposed to capture low frequency vibrations. However, this device uses contacting parts that increase the mechanical damping and it is capable of generating few nano-watts.

In this context, NEHSTech aimed at pursuing the following main general objectives:

1. to propose and validate innovative solutions to address miniaturization issues for electrostatic and electromagnetic motion-to-power transduction mechanisms;

2. to build an test novel versatile high power density and wide bandwidth VEHs at micro-scale (electrostatic and/or electromagnetic as second choice);

3. to advance the knowledge of nonlinear multi-mass oscillators applied to energy harvesting and disseminate the results to international level;

4. to train the fellow researcher on complementary expertise and skills such as: knowledge of the French language; knowledge of silicon-based MEMS VEHs (in particular, electrostatic generators based on comb capacitor) modelling, design, fabrication (lithography, DRIE etching, anodic bonding), characterization and (AFM, SEM microscopy). In addition, objective for the fellow to work in total autonomy, achieving research maturity and capability of academic networking.

Description of the main S&T results/foregrounds

NEHTech project has achieved the following S&T results/foreground:

• theoretical models of harvesting systems electro-dynamic with multiple-mass and nonlinear suspension were derived. These analytical models give new insight on the nonlinear behaviour with reference to energy generation on centimeter and sub-millimeter scale. An ad-hoc numerical simulation tool was developed by the fellow researcher with MATLAB code in order to study the behaviour of electromagnetic bistable VEHs and MEMS electrostatic gap-closing comb VEHs. The main outcome from these results is the possibility to predict the operation of nonlinear generators with different conversion technologies and/or physical characteristics.

• a centimetre-scaled electromagnetic harvester with buckled supporting steel beams was designed and built during the first half of the project (figure 1). The scope was to explore and understand its dynamical behaviour and the harvesting efficiency when the system operates in the bistable regime. An enhancement in the bandwidth response of a factor of 2.5 along with an and higher generated power off its natural resonance has been shown when compared to the monostable regime.

• an electrostatic silicon-based MEMS VEH, previously fabricated by ESIEE group (Guillemet and Basset) combined with an aluminium base supported by buckled steel beams was tested under different sources on vibrations: noise, harmonic sweeping, vibrations acquired from real world (bridges, vehicles). Despite the fact that the converter itself was designed to resonate at 162 Hz, the use of the bistable stage as mechanical exciter allowed a gain factor of hundred when converting low−frequency vibrations (20−40 Hz) into electrical energy. These results were presented to the conference IEEE MEMS 2013 (F. Cottone, et al., "Bistable multiple-mass elecrostatic generator for low-frequency vibration energy harvesting," in IEEE MEMS 2013 Conference).

• a novel prototype of silicon MEMS VEH based on electrostatic gap-closing comb was implemented and tested. This prototype is based on multiple mass velocity amplification effect combined to bistable springs and elastic stoppers. The bistability of the silicon mass is controlled by the bias voltage that operates a softening effect to the effective stiffness of the mechanical suspensions. The elastic stoppers serve to preserve kinetic energy and increase the velocity the proof mass and micro-ball during impacts. A power density of around 20µW/cm3 was demonstrated by the MEMS harvester at 0.3-g acceleration at very low frequency (10-60 Hz). A power gain up to a factor 5 has been shown with a -3db bandwidth of 50 Hz with respect to existing linear MEMS generators with single mass.

• measures of mechanical/electrical Power Spectral Density (PSD) were performed for different resistive loads and under a variety of realistic vibration inputs (random noise, sinusoidal, impulsive etc.). Comparative tests versus existing VEHs were also accomplished.


Illustrations of the initial prototypes for the concept investigation are shown in figure 1 and 2 in the attached pdf file. Images of the final NEHSTech prototype are not presented because the results are expected to be published in the near future in conferences and peer review journal papers.


In conclusion, the effectiveness of the proposed concept for vibration energy harvesting at micro scale has been proven. In particular, the capability of operating at very low frequency (10-60 Hz) for a MEMS harvester makes this technology potentially suitable for important applications such as peacemaker powering from human movements or self-powered wireless bridge monitoring. On the other hand, this first prototype was not yet fully optimized and further evolution is envisaged to improve its power performance. As an example, it is foreseen that this device will be using electrects materials so to eliminate the need for an external bias voltage.


The potential impact and the main dissemination activities and exploitation of results

So far, linear MEMS harvesters have never shown to be capable of generating usable power in the low frequency region (10-100 Hz) because of their high natural frequency (usually going from 1kHz to several kHz). The capability demonstrated by the NEHSTech technology of more than 20µ/cm3 is comparable with that of a lithium battery operating for 1 years. The advantage for the MEMS VEH is the possibility to work for almost unlimited time. Although in the absolute, the power result may appear not impressive, it constitutes only the validation of the concept and further optimization of the technology are envisaged. Nevertheless, the NEHSTech technology paves the way for future applications were self-powered wireless sensors are employed and were real vibrations have the main part of kinetic energy located at low frequencies. Some relevant examples are mobile devices, peacemakers, RFID active tags powered by human motion and bridge/building monitoring.

All the research progresses have been communicated during the whole duration of the project at informal meeting, schools, national and international conferences and workshops. Past and future publications and dissemination activities are listed in the following section.

Innovative aspects and variants of the concept that have been not presented in public because are envisaged for patent application. This task was delayed over the planned milestone because of lack in time due to the continuous evolution of the initial concept related to problematic aspects of the fabrication (structural aspects and geometry of the MEMS, dicing, dynamic principle and form factor).

In conclusion, all critical training objectives have been achieved during the first phase (first half of the project). In the second phase (the second half), a retard a few months in the fabrication of MEMS prototype was encountered with respect to the planning, but the key results of the project have been accomplished.



Project public website: http://www.esiee.fr/~bassetp/nehstech.html