Due to decreasing power requirements for wireless networks and other microwatt devices, energy harvesting from ambient vibrations has become a realized power source. Piezoelectric cantilevers are a viable option as the transducer element for converting mechanical to electrical energy. Being able to accurately model piezoelectric cantilevers is important in designing efficient converters needed in the power management circuitry. In this thesis a method is outlined that enables the modeling of the physical behaviour of piezoelectric cantilevers with an equivalent circuit model comprised of RLC circuits.
A MatLab model of piezoelectric cantilevers is developed and verified via literature comparisons. The equivalent circuit model, which is particularly important in electrical engineering of power management circuity, is demonstrated with a proof-of-concept harvester unit design. In the process and for a fraction of the traditional cost, the accuracy and usability of a low-cost vibration shaker is shown.
The design is tested on two real-world applications. The first application is a transformer that has regular vibrations with low harmonic content. The second application are industrial fans with high harmonic content vibrations. The results of both applications show that energy harvesting is possible with a simplistic approach as presented. Further it is shown that high harmonic content, such as was found in the fans, can interfere with energy harvesting in both a constructive and destructive manner leading to further analysis. These findings have implications in designing vibrational energy harvester units as discussed.
| Identifer | oai:union.ndltd.org:MANITOBA/oai:mspace.lib.umanitoba.ca:1993/23934 |
| Date | 02 September 2014 |
| Creators | Pielahn, Mathias |
| Contributors | Filizadeh, Shaahin (Electrical and Computer Engineering) Shafai, Cyrus (Electrical and Computer Engineering), Gole, Aniruddha (Electrical and Computer Engineering) Wu, Nan (Mechanical Engineering) |
| Source Sets | University of Manitoba Canada |
| Detected Language | English |
Page generated in 0.002 seconds