Unsteady aerodynamics, reduced-order modelling, and predictive control in linear and nonlinear aeroelasticity with arbitrary kinematics
Unsteady aerodynamics, reduced-order modelling, and predictive control in linear and nonlinear aeroelasticity with arbitrary kinematics
复制标题
任意运动学线性和非线性气动弹性的非定常空气动力学、降阶建模和预测控制
DOI:
--
复制
发表时间:
2015
期刊:
影响因子:
--
通讯作者:
R. Simpson
中科院分区:
文献类型:
--
作者:
R. Simpson
This thesis concerns the development of computational methods for efficient flexible-aircraft flight dynamics analyses. An argument is made for a hierarchy of methods that provide predictive capability for loads and stability analyses, and the ability to create low-order dynamic models for control system synthesis. The proposed aeroelastic models are formulated using three-dimensional unsteady aerodynamics in the form of an unsteady vortex-lattice method developed to model the relatively complex kinematics inherent in flexible-aircraft dynamics, and in particular the unsteady induced drag. No assumptions are made relating to the kinematics of the fluid-structure interface (inputs) and use of the three-dimensional Joukowski relation naturally resolves all components of the unsteady aerodynamic forcing (outputs). A consistent linearization of this method about an arbitrary reference state yields nondimensional (independent of free-stream dynamic pressure) discrete-time state-space models that resolve frequencies up to a spatio-temporal Nyquist limit defined by the wake discretization, and have a convenient form for coupling with structural dynamics models. Aircraft structural components are modelled using a geometrically-exact composite beam formulation, and, additionally, in the case of linear dynamics, a generic modal description. The latter allows the linear aerodynamics to be expressed in a reduced set of inputs and outputs, thus obtaining a time-domain alternative to the classical frequency-domain-based doublet-lattice method. The models modified for these modal degrees-of-freedom are shown to be amenable to balanced realization and truncation, and are verified in flutter analyses where only 10-100 balanced states are required (compared to 1000-10,000 physical states) for converged results. Finally, predictive controllers and linear-quadratic regulators are synthesized using reduced-order aeroelastic models, and are applied in nonlinear simulations for gust-load alleviation.
影响因子:
2.5
作者:
H. Hesse;R. Palacios
通讯作者:
H. Hesse;R. Palacios
影响因子:
4.7
作者:
Hesse, Henrik;Palacios, Rafael
通讯作者:
Palacios, Rafael