System Identification of Flexible Aircraft in Time Domain

Abstract

This paper proposes and demonstrates the applicability of the system identification methodology in time domain to include the effects of structural motion on the flight dynamics of an aircraft treated as an elastic body. For this end, flight tests using a highperformance sailplane equipped with special flight test instrumentation were performed. The structural motion is represented in terms of those normal modes which influence and are influenced by the rigid-body response. The coupling between rigid-body and structural motion is obtained through aerodynamic forces in terms of generalized stability and control derivatives, which are estimated using the output error method in the time domain. The conventional rigid-body stability and control derivatives as well as the deflections of the normal modes at the measurement points are estimated. The model dynamics incorporate two anti-symmetric and three symmetric normal modes. A validation analysis shows the improvements obtained by using the integrated model when compared to the traditional rigid-body approach. Nomenclature= Accelerations at measuring point , m/s 2 = Wingspan, m ̅ = Mean aerodynamic chord, m = Aerodynamic derivative representing the effect of the variable on the coefficient = Aerodynamic bias in the coefficient = Moment coefficients = Force coefficients along body axes = Lift and drag coefficients = Aerodynamic derivative representing the effect of the variable on the -th generalized load = Aerodynamic bias derivative of the -th generalized load ⃗ = Structural displacement, m = Drag efficiency (Oswald's) coefficient , = Acceleration due to the gravity, acceleration due to gravity at sea level, m/s 2 = Geometric altitude, m = Moments of inertia, kg.m 2 = Products of inertia, kg.m 2 = Aerodynamic moments, N.m = Aircraft mass, kg = -th structural modal mass, kg.m 2 = Number of flexible modes in the model = Angular rates, rad/s = Static pressure, Pa ̅ = Dynamic pressure, Pa = Generalized load acting on the -th structural mode, N.m = Air constant, J/(kg.K) 1 Guest scientist, Institute of Flight Systems, Lilienthalplatz 7, 38108 Braunschweig, Germany, on leave from DCTA -Institute of Research and Flight Tests, Brazil, epd-t@ipev.cta.br. 2 = Distance between horizontal tail aerodynamic center and the aircraft center-of-gravity, m , = Wing reference area, horizontal tail area, m 2 = Time, s = Static air temperature, K = Components of the velocity vector, m/s = True airspeed, m/s = Position of a point along the aircraft structure, m = Aerodynamic forces along the body axes, N , = Angle of attack, angle of sideslip, deg = Control deflections (ailerons, elevator, flaps, rudder), deg = Reference flaps position, deg = Position error, Pa = Pitot error, Pa = Measured strain at point = Bias of the strain measurement at point = Effect of the unitary -th structural mode displacement on the strain at point ⃗⃗⃗⃗ = Shape function of the -th structural eigenmode, m = Modal deflection at measuring point due to the -th modal displacement, m = Aspect ratio = Displacement of -...

Attenuation of Aircraft Flexible Modes during Maneuvering Flight

“…An extension of the results is presented in Ref. [14], where a model containing three symmetric and two anti-symmetric structural modes is identified and validated. Both Refs.…”

“…Both Refs. [13] and [14] use a model based on the one developed at Ref. [4], where the coupling between rigid body and structural motion is obtained through generalized derivatives, which seems to be a natural extension of the well-established approach adopted in the system identification of rigid body aerodynamic models.…”

Nonlinear aeroelastic modeling of aircraft using support vector machine method

“…In the time domain, the simplest method for the system identification of elastic aircraft is the equivalent stability and control derivatives in which the rigid aircraft parameter estimation techniques are applied to flight data of elastic aircraft (Giordano de Oliveira Silva and Moennich, 2012). However, because of the linearity of derivatives, this method is not successful in dealing with the complex flight dynamics of the elastic aircraft.…”

Identification of Aeroelastic Models for the X-56A Longitudinal Dynamics Using Multisine Inputs and Output Error in the Frequency Domain

“…This simplified model is also used in Reference [13], where the filter-error approach is used to identify models for the SR-71 aircraft from measured time histories. In Reference [14], the full Waszak-Schmidt model is used to identify a model of a high-performance glider using an output-error analysis in the time domain. In References [15,16], flight data from frequency sweeps and multisine inputs are used with output-only modal identification and frequency response estimation to identify aeroelastic models of a small, unmanned aircraft.…”