Uncontrolled aerobatic tumble manoeuvre mechanics

Abstract

Despite advancements in aerobatic aircraft, the issues concerning aerobatic flight safety remain largely unresolved with incidents occurring at airshows, during training and during private flights, even with stable, predictable and reliable aircraft. The likelihood of aerobatic incidents has increased as aircraft are now capable of being flown in even more dangerous, low energy scenarios. Spins or tumble manoeuvres are typically the origin of loss-of-control accidents. Such scenarios occur generally in manoeuvres that result in low energy states. This dissertation aims to understand the mechanics behind a low energy aerobatic manoeuvre. The positive 𝑔 tumble manoeuvre was selected as the manoeuvre to analyse, as it is one of the most difficult aerobatic manoeuvres to recover from. The Extra 330SC was selected as the model aircraft as it is one of the most popular aerobatic aircraft across all pilot skill levels. Additionally, the Extra Aircraft Corporation provided more information about the aircraft than any other aerobatic aircraft manufacturer. Methods to develop mass, inertial and aerodynamic data were explored. The aerodynamic data was extended to high angles of attack. Aerodynamic models for the wing, fuselage, horizontal stabiliser and propeller were considered. Generated mass and inertial data were compared to reference data and demonstrated good correlation. Static aerodynamic data was compared to published experimental data (where possible). Good agreement was shown between the generated data and test data. Considerations for surfaces exposed to propeller slipstream and rotational dynamics were considered. No literature was available for direct comparison of these considerations. A non-linear three degree of freedom model was developed to simulate the tumble manoeuvre. This was achieved by simplifying the equations of motion in the velocity axes. The velocity axes were appropriate in determining the flight path of the manoeuvre along with all affiliated parameters. The aircraft body axes were utilised in determining the rotational parameters during the tumble manoeuvre. Force components at high angles of attack were modelled in both the velocity and body axes. The current model does not include any lateral-directional forces or moments, assuming the manoeuvre is purely in the longitudinal plane. A tumble manoeuvre is possible for a very specific aircraft configuration. An upgraded engine and propeller slipstream effects are essential in performing the tumble manoeuvre. It was shown the entire manoeuvre could be completed in a horizontal distance of ≈ 50 m and a vertical distance of ≈ 15 m for manoeuvre entry velocities of 25 m/s, 30 m/s, 35 m/s and 40 m/s. Each of the minimum radius tumble manoeuvres occurred at the most rearward centre of mass location. An entry velocity of 30 m/s provided the smallest possible radius tumble. The entry to the manoeuvre requires large decelerations and normal forces that were beyond the lift limits of the wing. Both these requirements were met by orientating the aircraft at large pitch angles and subsequently large angles of attack. The large angles of attack resulted in large decelerating forces and orientated various aircraft force components, other than iii the wing, in the lift direction. Significant energy loses are seen through the manoeuvre resulting in large losses in airspeed and ultimately altitude. The risks associated with a tumble manoeuvre could be minimised by adjusting the normal forces and moments acting on the aircraft. By altering the propeller blade pitch angle, the normal forces and subsequently the moments from the propeller could be changed, altering the overall normal forces and moments acting on the aircraft. This resulted in a slightly larger overall manoeuvre but illustrated increased velocity values throughout the manoeuvre. A form of validation of the results was performed by comparing model results of tumble manoeuvres that were generated through a patchwork of images compiled from video footage of the manoeuvre. Good correlation was shown between the model data and video footage.