An exploration of experimental and numerical approaches to design of a ducted helical air turbine

Abstract

The proposed device is a ducted air turbine with a 2-bladed helical rotor. To attempt to characterize the performance of the device, as well as establish a methodology for further experimentation, a campaign of testing was undertaken. The campaign described in the work culminates in a comparative study between a simulated experiment (undertaken using Computational Fluid Dynamics) and a physical experiment (undertaken with a dynamometer, and a duct system connected to a flow meter and centrifugal air pump). The simulated experiment has been transformed using the similarity laws, and the resulting data has been used to predict the performance of the physical experiment by interpolating to the setpoints observed in the physical experiment. These setpoints are defined (in both the simulated and physical experiments) by experimental input variables: bulk flow velocity of the fluid, and rotational velocity of the rotor. Each successful experiment produces experimental output variables: braking torque applied to the rotor, and pressure drop across the duct section which encloses the rotor. A direct comparison of simulated and physical performance data through the use of nondimensional coefficients demonstrates good agreement between the two experiments, though some discrepancy in torque has been identified. The degree of agreement suggests that this implementation of CFD and the similarity laws would be a good basis for future analysis of turbine performance.