A physical prototype manufactured in-house (Loughborough lab) demonstrates the investigated properties. Two aluminium housings are used to place the rigid magnets, and a plastic driving rod connects them to provide a rectilinear driveway for the oscillating magnet. An additional aluminium base is used to mount the harvester body on a vibration shaker.

Figure 17. Construction of a physical prototype and calibration testing in a vibration shaker

The shaker tests provided data for identifying the realised stiffness and damping properties of the harvester, leading to refined computational models. The harvester is then mounted on a hub that is attached on a spinning shaft. A DAQ board is also attached on the harvester body to allow measurements of the induced voltage.

Figure 18: (a) Frequency response to translational vibrations utilised for system identification; and (b) mounting of the rotational harvester on a spinning shaft with on-board DAQ

Initial testing of the rotational harvester included increasing frequency ramps of the mounting shaft. Multiple resonant peaks were expected from our modelling results, due to direct forcing of the harvester both in mono- and bi-stable configuration. The experimentally observed voltage verified our modelling predictions, showing multiple resonance zones inherent to the harvester’s response. These experimental observations build confidence on the potential of this concept to operate across broad frequency ranges that are commonly met in automotive powertrains.