Structural optimisation through material selections for multi-cantilevered vibration electromagnetic energy harvesters

• Main Authors: Mohamad Faiz Foong, Muhammad Faruq Foong, Thein, Chung Ket, Yurchenko, Daniil
• Format: Article
• Published: Elsevier Ltd 2022
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://eprints.utm.my/103137/; http://dx.doi.org/10.1016/j.ymssp.2021.108044

This study investigates the structural optimisation of a multi-cantilevered electromagnetic anti-phase energy harvester by considering the optimum material choice. The mathematical model for the harvester was first developed, displaying an excellent correlation when compared with the experimental results. Afterwards, the anti-phase harvester was structurally optimised under a defined set of constraints while only considering a single material for all cantilever beams. Here, three materials with low damping capacity were considered. It was found that if the beam thickness was unchanged, the optimisation for certain materials would not converge due to high stress levels exceeding the fatigue safety factor of 80.0%. The safety factor was implemented in the optimisation to ensure the device's durability. However, the unsuccessful material exhibited the lowest damping capacity among other materials. Hence, a mixed material approach was attempted which produced an optimum power output of 119.6 mW and a 1.76 V voltage output under a base acceleration input of 0.1 g and a practical volume constraint of 600 cm3. This corresponds to a 33.3% increase in power output when compared to the single material harvester. Further analysis demonstrated that if the fatigue safety factor was lowered to 60%, the optimised power output would drop by 26.5% to 87.9 mW whereas a slight increase in voltage was recorded. Finally, the optimum material for the masses on the harvester was examined, suggesting that a high-density material must be used for the beam clamp and the proof masses on the magnet beams whereas the mass on the support beam must be minimized.