Development of a pendulum-based device to harvest energy from large amplitude, low-frequency oscillations

Abstract

The rapid growth of the Internet of Things (IoT), wearable electronics, and remote sensing technologies has created strong demand for sustainable, self-powered devices. Vibration energy harvesting aims to meet this need by converting ambient mechanical energy into electricity. However, traditional spring–mass harvesters are limited by static deformation (vertical type), friction losses (horizontal type), restricted motion range, and narrow bandwidths, while conventional pendulum harvesters typically require large motion clearance, reducing compactness and power density. To address these challenges, this work develops a compact pendulum-based electromagnetic energy harvester (EMEH) integrating spiral torsional springs and a dual Halbach magnet array. The spiral springs serve as motion limiters, maintain axial compactness, and allow large angular deflections, whereas the Halbach array produces intensified unidirectional magnetic fields to enhance electromagnetic coupling. A lumped-parameter model is formulated to describe the coupled electromechanical dynamics. Spiral springs are designed and 3D-printed in PLA, and their stiffness is characterized through finite-element analysis, static torque–angle testing, and natural- frequency measurements. Two coil arrangements for the EMEH are examined, and their electromagnetic transduction factors are obtained from COMSOL magnetic-flux simulations. Using the computed transduction factors and the dynamic model, free-voltage responses across various load resistances are generated via numerical simulation and compared with experimental measurements. Results show that the second coil configuration yields closer agreement between simulated and measured responses, indicating stronger electromagnetic coupling and improved model fidelity. The energy harvesting performance of the proposed EMEH is further examined through systematic free-response and base-excitation analyses across five mechanical configurations, with different stiffnesses and mounting strategies. Performance metrics including harvested electrical energy, peak power, average power, and energy conversion efficiency are evaluated over a wide range of load resistances and excitation amplitudes. Harmonic and swept-frequency numerical simulations are employed to investigate the system response under sinusoidal base excitation, revealing the trade-offs between stiffness-induced frequency tuning, motion regulation, bandwidth variation, and electrical power output.