Files
Abstract
External aerodynamic simulation of a rotor system is far more challenging compared to simulations involving fixed wing aircraft. This is because the surrounding flow field itself is far more complex. The rotating blades of the rotor system encounters its tip vortex and as well as the tip vortices of other blade in a phenomenon known as Blade Vortex Interaction (BVI). These complexities make the modeling of a rotor system and accurate predictions of rotorcraft performance challenging. Rotary wing Micro Aerial Vehicles (MAV) are such systems that have unique capabilities of vertical takeoff and lift (VTOL), hover, and low-speed flight. Quadcopters are such rotary-wing MAV that involve multi-rotor systems. Given that the modeling of a single rotor system is complex, modeling of a system with multiple rotors is even more challenging due to the rotor - rotor and rotor - fuselage interactions. This thesis explores and evaluates the computational modeling of a commercial quadcopter in hover using Computational Fluid Dynamics (CFD) simulations. A high fidelity CFD simulation methodology is first developed for an isolated rotor through investigating the effects of various simulation parameters like meshing schemes, turbulence modeling, and flow-physics setups on the veracity of the CFD predictions. Rotorcraft performance parameters such as the thrust coefficient, torque coefficient, Figure of Merit, and blade surface pressure coefficients are validated against and correlated to experimental data obtained using the Caradonna and Tung rotor. Proper resolution of the rotor wake region to prevent any grid related dissipation is observed to be critical for a reliable CFD simulation. This simulation methodology is further validated by comparing CFD simulations of an isolated rotor of the quadcopter (DJI Phantom- 3) carried out against available experimental results. This double verification approach enables the developing and testing of a CFD methodology for rotorcraft flow-field simulations. Unsteady Reynolds-Averaged Navier-Stokes (URANS) and Improved Delayed Detached Eddy Simulation (IDDES) models are used in the case of DJI Phantom-3 quadcopter to predict its performance and of its individual components. Limitations of URANS in accurate prediction of rotor-rotor and rotor-fuselage interactions are studied. Differences in predictions of commercial CFD code StarCCM+ against OVERFLOW code is studied for quadcopter applications.