Defining robust transition and re-transition procedures for unmanned fixed-wing vtol aircraft

Fixed-Wing Vertical Take-Off and Landing (VTOL) aircraft are widely discussed due to their diverse applicability. For such vehicles, hover flight is commonly enabled by dedicated or partially dedicated motor-propeller units, whereas in fixed-wing flight the lift is produced from aerodynamic surfaces. The so-called transition and retransition (also back-transition) of these aircraft involves the deactivation and activation of the hover propulsion system in a specific airspeed region above stall speed. The commonly available implementation of the transition involves a ramp-up of the throttle, where upon exceeding a certain airspeed, the hover propulsion system is shut down in a fixed time frame. For the retransition, the thrust is taken away and the propulsion system is activated when the aircraft decelerates to a specific airspeed value. The activation of the propulsion system during retransition as opposed to before take-off occurs at significantly different airflow conditions and hence loads on the propellers. This has the consequence that turning on one or more motors can fail. If unaccounted for, the aircraft may become uncontrollable if it decelerates below the stall speed. In this work we propose a transition and retransition procedure for safety-critical applications that considers and guarantees safe operation in such events. Furthermore, by design the procedure prohibits the aircraft reaching high airspeed with an active hover propulsion system or stalling with a not fully activated one. We derive a minimum set of controller functionalities necessary to execute such a procedure. Without loss of generality, the suggested solution can be applied for both visual and beyond visual line of sight. The procedure is implemented and tested with an existing tiltrotor fixed-wing VTOL aircraft and controller architecture. This paper presents simulation results of the above mentioned procedure in the nominal case, where no component faults occur. The robustness of the procedure in the case of failures is demonstrated in simulation.