Power harvesting for railroad track health monitoring using piezoelectric and inductive devices

One of the most limiting factors for distributed sensor networks used for railroad track health monitoring applications is the lack of a long-term, low-maintenance power supply. Most existing systems still require a change of battery, and remoteness of location and low frequency of maintenance can limit their practical deployment. In this paper we describe an investigation of two principal methods for harvesting mechanical power from passing railcars in order to supply electrical power to remote networks of sensors. We first considered an inductive voice coil device directly driven by vertical rail displacement. We then considered a piezoelectric device that is attached to the bottom of the rail and is driven by the longitudinal strain produced by rail bending due to passing railcars. Theoretical models of the behavior of these devices were integrated with an analytical model of rail track deflection to perform numerical simulations of both of these power scavenging techniques. Lab and field tests were also performed to validate the simulation results. Resulting values of average power production show promise for scavenging near the targeted level of 1 _mW, and the field data matched well with the simulations.