A Study on Flexible Aircraft Flight Dynamics considering Gust

Abstract

In recent years, staying aloft for extended periods of time, aircraft wing has become slender to meet the needs to be light and increase aerodynamic efficiency. Due to the wing with large aspect ratio, the flexible wings might be deformed even at trim. The aircraft flight dynamics and structure response is more affected by the gust which might occur during flight. In order to consider the coupled effect of flight dynamics and aircraft structure flexibility, floating frame of reference formulation is used to compose the flexible aircraft equations of motion. This aircraft structure model is combined with finite-state unsteady subsonic aerodynamics considering the control surfaces. The discrete gust model is incorporated into aerodynamic model. The gust load is one of the unexpected natural phenomena that affects the aircraft flight dynamics. The gust load analysis is important process to evaluate aircraft flight performance and structure integrity. The quasi-static approach of the gust loads analysis is well established for the conventional aircraft to evaluate the increased load factor due to the gust. This approach is based on the assumptions and empirical data such as rigid-aircraft, level-flight, constant flight condition and critical gust length, so on. This approach enables low-fidelity calculations of the increased load factor without solving the equations of motion. In recent years, the aircraft wing has become slender to increase the aerodynamics efficiency. If the aircraft has slender wing, which can be deformed, the flight dynamics and the structure responses are more affected by the gust. Due to the aircraft flexibility, the quasi-static approach is not appropriate to evaluate gust load. To evaluate high-fidelity calculation of the gust effect, the transient analysis might be performed. In this study, the transient analysis of the gust load using the six-degrees-of-freedom flexible aircraft is proposed and compared with the quasi-static approach. From the comparisons, it can be seen that the gust loads of transient approach is more precision than quasi-static approach. Moreover, the gust load under aircraft maneuvering, which cannot be analyzed quasi-static approach, can be evaluated through the proposed approach. To evaluate the aircraft structure reliability and flight performance, the flight test simulations are performed. The flight test data is analyzed for the force estimations to improve the aircraft characteristics. The gust effect is one of the important sources in the force estimation. Therefore, gust effect needs to be considered along with the aerodynamic force, gravity and thrust. The gust generates the unexpected force which changes the flight path and structure responses from the nominal flight path. Two more aspects should be considered in the force estimation using the flight test data. First, the limitations on the equipment installation due to its location and weight restrict the amount of response data. Secondly, as the aircraft wing gets slender to increase the aerodynamic efficiency, the gust effect becomes complicated because of the coupled effect of flight dynamics and aircraft structure flexibility. In this study, we focus on the force estimation for the flexible aircraft by using optimization methods. Flexible aircraft analyses are performed with an aircraft model based on flexible multibody dynamics where a gust model is incorporated into the aerodynamic model. The limitations on the flight data are also considered in the model construction. The gust parameters, which generate the same response to the reference data, are identified in the optimization process. For considering the many case of flight conditions, the computation time should be efficiency. Reduced order modeling (ROM) techniques have also been adapted to increase computational efficiency. The ROM based on proper orthogonal decomposition (POD) is presented. The computation time, on the other hand, might be reduced to 40~50% comparing to the full model analysis because reduced degrees-of-freedom (d.o.f) iteratively should be recovered to full d.o.f. The ROM which is based on POD is improved by using artificial neural network (NN). By constructing surrogate model, the iteratively recover process to calculate force matrix can be removed. The ROM with NN computation time is reduced to under 10% comparing the full analysis.