The problem we are tackling is fundamentally that of energy conversion. The water in the stream contains kinetic energy. The turbine takes this into the mechanical domain in the form of rotation. The speed is increased through the belt and gear-train to ensure the brushless DC electric motor generates a sufficient voltage.
The energy now enters the electrical domain. Despite being called a DC motor, it produces a three-phase output that we rectify and process to interface with the battery. The unit includes over-voltage protection and maximum power point tracking to optimize charge rate while keeping the battery healthy to ensure long term operation is possible.
The waterproof container at the top contains the power electronics mounted to a 3d printed frame that makes accessing all components very easy.
A custom made printed circuit board houses the rectifyer, boost converter, MPPT (maximum power point tracking) functionality and a ATMega328P microncontroller.
By choosing to design a dedicated propeller, we were able to optimize the system for our operating conditions by using computational fluid dynamics. Using the SOLIDWORKS Flow Simulation package, a sweep of blade count, flow velocity, and blade pitch enabled us to choose an optimal combination of values. Given a 1.4 m/s flow and the NACA2403 foil, a turbine blade count of 12 was found to produce the most power.
We have tested the prototype in multiple streams and conditions to ensure it operates to expectation in a variety of situations. For sufficient water depth and flow velocity, the system can produce significantly more than the 5W target.
The data collected from Wreck Beach and Seymour River displays how efficiency rapidly decreases when the water depth is too low and when the water velocity is too slow respectively.