Application of aircraft's flight testing techniques to the aerodynamic characterization of power kites

Abstract

This thesis has developed an experimental methodology for the flight testing and data analysis of power kites applied to Airborne Wind Energy Systems (AWES). In particular, the Estimation Before Modeling technique, a well-known method in the aerospace industry for the aerodynamic characterization of an aircraft using real flight data, has been adapted for tethered aircraft. The developed methodology has two main building blocks: (i) an experimental setup to record experimental data during the flight testing, and (ii) a Flight Path Reconstruction algorithm to estimate the state of the system from the experimental data. From them, the aerodynamic characteristics of two types of kites were investigated. The proposed experimental setup was designed to be low cost, portable and easily adaptable to both, rigid and semi-rigid kites. It is composed of an instrumented kite representative of the ones used in AWES, an instrumented control bar, a ground computer and a wind station. Whenever it was possible, commercial off the shelf components have been used, including low cost openhardware sensors based on the PixHawk platform. However, after the first flight tests were conducted and the obtained results were discussed, high precision sensors were also included. The Flight Path Reconstruction (FPR) algorithm for tethered aircraft is based on an Extended Kalman Filter (EKF). In addition to the standard set of estimated state variables (ie. Euler angles, position or ground speed), the algorithm also provides the aerodynamic torque and forces upon the kite as well as the tether tensions and wind velocity vector. The EBM technique, and the FPR algorithm have been used to identify the aerodynamic characteristics of both, four-line Leading Edge Inflatable (LEI) kites and two-line Rigid Frame Delta (RFD) kites. Quantitative and qualitative results have been obtained. Albeit both types of kites exhibited very high AoA during the flight, some significant differences were found. In particular, the estimated lift coefficient of the LEI kite showed a behavior identified with a post-stall condition, while the RFD showed a pre-stall behavior with a lower AoA and a positive relation between the lift coefficient and the kite AoA. The presented experimental methodology can be of great interest for AWE industry as it helps to improve modeling of tethered aircraft, leading to more accurate performance figures which may increase investors interest in the technology. Moreover, flight testing methodologies and experimental data analysis are of great interest for benchmarking AWES performances, contributing to de-risk their development process and providing better tools for AWE "best concept" identification. Finally, as a sub-product of the presented methodology, the FPR algorithm can be used as a validated state estimator of the tethered aircraft, which is a key element of a closed loop flight control system.