Modelling and Characterization of a High-Efficiency, Cm-Scale and Low Velocity Airflow-Driven Harvester for Autonomous Wireless Sensor Nodes

This paper reports the design, simulation, fabrication and performances of a centimeter-scale $(\emptyset=35\mathrm{m}\mathrm{m})$ airflow-driven harvester for autonomous Wireless Sensor Nodes (WSN). We present a model-based design tool implemented in Matlab-Simulink, which takes both computational fluid dynamics and electromagnetic fmite element simulations as inputs and we compare the simulation results with measurements for various air velocities. The harvester has a cut-in speed of 2 $\mathrm{m}.\mathrm{s}^{-1}$ and it is particularly efficient in the low airflow environments since its end-to-end efficiency ranges from 10.5% to 23.9% and its maximum output power from 200 $\mu \mathrm{W}\mathrm{t}\mathrm{o}3.7\mathrm{m}\mathrm{W}$ at 1.5 $\mathrm{m}.\mathrm{s}^{-1}$ and 3 $\mathrm{m}.\mathrm{s}^{-1}$ respectively. The propeller alone has a mechanical power coefficient ranging from 19.1% to 34% at 1.5 $\mathrm{m}.\mathrm{s}^{-1}$ and 3 $\mathrm{m}.\mathrm{s}^{-1}$ respectively. Furthermore, in the cm-scale and low airflow velocity ranges, the proposed harvester without shroud outperforms the state of the art in terms of power density and end-to-end efficiency (23.9% at 3 $\mathrm{m}.\mathrm{s}^{-1}$, 28% at 5 $\mathrm{m}.\mathrm{s}^{-1}$) and it still exhibits one of the highest performances with its shroud.