Hydrodynamic Optimization of Ducted Hydrokinetic Turbine

It has long been hypothesized that a duct can accelerate and condition the fluid flow passing the hydrokinetic turbine and improve overall energy extraction efficiency. To investigate this problem, we explore the optimal design of a ducted hydrokinetic turbine to maximize hydrodynamic efficiency in this work. Our method relies on combining both gradient-free and gradient-based optimizations. We first conduct gradient-free Bayesian optimization in conjunction with a low-fidelity steady RANS varying a few important design parameters. The optimal turbine geometry obtained is then used as the baseline design for subsequent gradient-based optimization using the adjoint method with substantially more design variables (representing both blade and duct geometries). The optimized geometry is finally evaluated through unsteady RANS simulations.