Flapping-wing UAVs, more specifically, ornithopters have been used recently in different applications. For example, in airports, birds tend to collide with airplanes during their take-off and landing causing damage to the planes and death to the birds. One way to solve this is to use bird-like robots to scare these birds away and prevent a collision. To have an ornithopter that can be controlled to do this, a dynamic model has to be developed. Since flapping ornithopters are physical systems that are governed by energy, the focus of this thesis is to develop a fully dynamic model in an energy-based manner. This work provides a systematic way of rigid-body modeling in the Port Hamiltonian framework based on the work of (27). The procedure is applied on a multi-body flapping-wing UAV. The final port Hamiltonian model is an open model that can be connected to other sub-systems such as the air and a controller subsystem. The connection between the preliminary model and the air resembles the aerodynamic contribution to the flapping-wing UAV, which is the second contribution of this work.
Flapping flight is a highly complicated mechanism exhibiting unsteady behavior. Scientists throughout the last century and till now have been studying the physics behind flying animals. They adopted aerodynamic models to predict their behavior. In most literature, they developed quasi-steady assumptions. These models ignore the unsteady wake effects and do not capture most of the unsteady phenomena of flapping. This thesis focuses on a realistic, unsteady aerodynamic model that takes into account most of the unsteady phenomenon of flapping. The model adopted in this work is the model by Delaurier (8) in 1993. The aerodynamic model is connected to the port hamiltonian dynamic model of flapping-wing UAVs. The model was validated and proven to be working through a set of four experiments. The first experiment is a time history to predict the generated aerodynamic forces. The other three experiments are sweeping over different flight parameters such as the pitch angle of the bird µa, flight speed U and flapping frequency f. An optimal value of flapping frequency that maximizes the thrust and generated positive lift, was observed through the 4th experiment. Interestingly, this value agrees with previous work that was performed on the same model used in this thesis, the Robird. With that being said, this thesis provides a fully dynamic model in the port Hamiltonian framework, that can be used as a plant for future control purposes.
To join the presentation via Microsoft Teams click here