High-fidelity environment coupling CFD and multibody dynamics for the study of flapping drone aerodynamics

For the past few years, insects and hummingbirds are inspiring a new generation of drones called Flapping Wing Micro Air Vehicles (FWMAVs). Flapping wings show excellent aerodynamic performance, adapt to very different wind conditions and allow unmatched aerobic maneuvers. These features could be valuable for flapping drones in various applications including rescue, military and civil missions. In the design of optimal flapping motions, simplified simulation tools are often used. Quasi-steady models estimate the aerodynamic performance of the wings and 6 degrees of freedom equations predict the resulting motion of the drone. While fast, those models result in different trajectories compared to real experiments. In addition, the models don’t provide details on flow physics. The present thesis aims to develop a high-fidelity, multi-physics environment that couples CFD to accurately compute the wing forces and a multibody dynamics system (MBS) to accurately predict drone trajectories. Different coupling strategies between CFD and MBS are first evaluated in light of recent publications. The rigid body dynamics libraries already implemented in the CFD software OpenFOAM are used. The rigid multibody system is defined with the forward dynamics formalism and solved with the projection method. The dynamics is time-integrated and subsequently, the CFD solves the laminar flow around the flapping drone to compute the generated aerodynamic forces. The CFD uses the overset meshing technique to move the grid according to the wing motion. The latter is imposed either through a direct approach, which drives the dynamics of the wings within the projection method or an indirect approach, which restraints the wing torques according to a PID controller. To evaluate the two strategies, a 2D system is defined with a flapping and pitching airfoil combined with a two degrees of freedom body (vertical and horizontal translations). The results demonstrate that the direct approach provides an efficient and flexible way to impose the flapping motion. The simulations allow to capture complex unsteady flow mechanisms and reveals the aerodynamic interaction between the body and the wing. The multibody environment is then extended to a 3D, six degrees of freedom point drone with elliptical wings. Simulations are performed for a few flapping motions in order to evaluate their influence on the drone trajectory. The present work provides an accurate environment to simulate the flight of flapping drones and investigate the complex flow phenomena manipulated by flapping wings. The presented work could then serve as a benchmark to highlight the limits of simplified aerodynamic models in different flight scenarios.