Hypersonic Aeroelastic Analysis Based On Parameterized Reduced-Order Model

Hypersonic vehicles generally refer to the flight vehicles flying in the atmosphere layer or trans-atmosphere layer at a flight Mach number above 5,which is an important direction in current and future developmet of flight vehicles.A hypersonic vehicle usually adopts a long,slender lifting body layout,and the body and wings are flexible due to minimum-weight restrictions.Due to complex interactions between flows,flight dynamics,structural dynamics,controllers,and propulsion system,the aeroelasticity and aerothermoelasticity of the hypersonic vehicle have been created.The accurate and efficient predictions of aerodynamics,aerothermodynamic loads,structural temperature distribution,thermal deformation and thermal stress as well as vibration response of thermal structural are the very important and challenging tasks.Based on the research status of hypersonic aeroelasticity at home and abroad and the engineering application background,aerothermoelasticity of the hypersonic wing is studied systemically under the severe aerodynamic heating environment in this work.The parameterized order reduction methods for hypersonic unsteady aerodynamics and the thermal mode reconstruction method under the influence of aerodynamic heating are established to explore an efficient and accurate hypersonic aerothermoelastic analysis method.At the same time,the hypersonic aeroservoelastic analysis based on parametric order reduction method of unsteady aerodynamics is studied.The main contribution of the dissertation are summarized as follows:(1)For a airfoil,hypersonic aeroelasticity is simulated by using CFD-based techniques,and the effects of angle of attack and flight altitude on flutter and limit cycle oscillation are analyzed.Based on CFD and computational thermo structural dynamics techniques(CTSD),the thermal modes of a three-dimensional wing with low aspect ratio are calculated.The effects of aerodynamic heating on structural vibration characteristics under different flight conditions are analyzed and aerothermoelastic simulation is carried out.(2)According to the quasi-steady nature of hypersonic flowfields,a parameterized reduced order method(ROM)for hypersonic unsteady aerodynamics is proposed.First,the quasi-steady nature of the flow was exploited to approximate the loads using steady-state CFD combined with corrections for unsteady effects using simple analytical expressions derived from piston theory.The computational cost of the steady-state CFD was minimized by using ROM techniques based on proper orthogonal decomposition(POD)and Kriging interpolation.Numerical results show that the proposed parametric ROM has advantages of hight efficiency and accuracy over a wide range of parameter space.(3)An adaptive POD method is proposed based on the concepts of Grassmann manifold and manifold geodesic line,and by combining least squares support vector machine(LS-SVM).The method of interpolation in a tangent space to a Grassmann manifold was used to generate a new POD modal matrix for the arbitrary operating point in the defined parameter space.LS-SVM was utilized to obtain the dynamic relationship between the applied excitation and the corresponding POD coefficients.Thus,a parametric ROM of unsteady aerodynamics is constructed.The aeroelastic simulation can be carried out by coupling the ROM with the structural equation.A comparing study with the direct CFD simulations indicated that the parametric ROM is demonstrated that the parametric ROM does efficiently predict aerodynamic behaviors,flutter boundary,and limit-cycle oscillations.(4)A thermal mode reconstruction method based on Grassmann manifold tangent space interpolation and Kriging interpolation is proposed.Selecting limited sample points in the defined flight parameters,modal data at each sample point are obtained by CFD and CTSD techniques.The Grassmann manifold tangent space interpolation method is used to obtain the modal matrix at any flight condition of the parameter space.The Kriging interpolation is used to establish the approximate relationship between natural frequencies and design variables.Thus,the corresponding thermal modal data can be obtained quickly and accurately in the parameter space.An efficient and accurate hypersonic aerothermoelastic analysis method is constructed by combining the above parameterized ROM for unsteady aerodynamics.In the parameter space,the method can accurately predict the fast aeroelastic response,and the efficiency is greatly improved compared with CFD/CTSD coupling method.(5)The proposed parameterized ROM of unsteady aerodynamic is applied to hypersonic aeroservoelastic analysis.An adaptive flutter suppression controller is designed by using active disturbance rejection control(ADRC).LS-SVM is embedded in classical ADRC to improve the control quality and robustness of the controller.SVM-ADRC flutter suppression controller has synthesized for an airfoil model of three degrees of freedom.Numerical results show that SVM-ADRC controller has better control performance in parameter space and can greatly increase the Mach number of flutter by comparing with classical ADRC controller.