Aero-structural Design and Optimisation of Tethered Composite Wings

Abstract

Airborne wind energy (AWE) is an emerging renewable energy technology that harnesses wind energy using tethered flying systems. The extra degrees of freedom allow these systems to harvest wind resources at altitudes currently unrealisable by conventional turbines. These flying devices, often resembling kites or drones, are typically divided into two classes. The first converts kinetic energy into electricity using onboard generators and transmits it to the ground via a conductive tether. The second class transfers aerodynamic forces via the tether to the ground, where the mechanical energy is converted into electrical energy using an electrical machine.As the tether’s length constrains the system, once the flying device reaches this tether length limit, some energy must be used to retract it back to its initial position. This cycle of traction and retraction is known as a pumping cycle. Therefore, AWE systems must be designed to maximise the harvesting or traction phase while minimising the retraction phase to ensure a net positive power output.
From the AWE system landscape, this thesis is based on tethered aircraft-style fixed-wing systems. Typically, such systems utilise composite structures owing to their high stiffness-toweight ratios. Designing these composite structures demands special attention due to their anisotropic nature, which results in complex load-deflection couplings. Here, a multi-disciplinary simulation framework for tethered composite aircraft wings is developed. The research focuses on methods used during the iterative phases of initial (conceptual and preliminary) design that are commonly employed in a spiral system engineering approach. The proposed framework integrates computational methods for the design of the aerodynamic A, bridle B, and structural S domains. The bridle is a system of segments of tether and pulleys that distribute the tether forces into the wing structure. The aerodynamic and structural domains are divided into 2D and 1D models, which are then integrated to determine the 3D response of the wing. A nonlinear vortex-lattice method (VLM) is utilised for the aerodynamic domain.
For the structural domain, an anisotropic 1D finite element (FE) model is developed that is coupled with a 2D FE sectional solver. In addition, methods are proposed that enable detailed topology optimisation. For tailless swept-wings, like those used by EnerKíte, the aero-structural-bridle interactions are crucial. The developed framework is used to investigate the impacts of different wing and bridle configurations to determine the sufficient level of fidelity required at the initial design phases. Typically, such aeroelastic phenomena are captured during detailed design stages wherein full 3D structural and aerodynamic simulations are employed. However, this mandates design knowledge typically unknown at the initial design stages. This motivates a multi-fidelity modelling approach to include these coupling effects while abstracting the composite ply level details during the design exploration. This is achieved by combining geometric discretisation approaches with lamination parameters. Thus, the framework aims to provide viable design options during the initial stages while considering aero-structural-bridle couplings.