In this video, Knight introduces research in his group.

A recent study presents a multifidelity computational fluid dynamics (CFD) framework for the design and optimization of ducted hydrokinetic turbines. These turbines harness energy from riverine and marine currents, and the ducted configuration enhances flow control and energy capture. The framework integrates three levels of fidelity: low-fidelity body-force models for rapid initial optimization, medium-fidelity blade element momentum theory coupled with CFD for capturing duct-rotor interactions, and high-fidelity rotating sliding mesh simulations for final validation.

High-performance computing was essential for executing the most detailed simulations, which were performed on the UMass UNITY cluster. This resource enabled the team to balance computational cost with accuracy across design stages, validating their models against experimental data and refining turbine performance predictions. The study demonstrates how strategic use of computational resources can accelerate renewable energy technology development.