A Volterra Kernel Reduced-Order Model Approach for Nonlinear Aeroelastic Analysis

Abstract

A procedure to develop CFD-based reduced order models (ROMs) which capture the essence of an aerodynamic system while reducing the complexity of the computational model is introduced. The Volterra-based ROM is obtained using the derivative of unsteady aerodynamic step-response. Transient responses and Gaussian responses were calculated using ROM and compared with the CFD solver for validation of ROM approach. An Eigensystem Realization Algorithm is used to convert ROM unsteady aerodynamics into the LTI state space model. A reduction in the cost of the realization of the ROM kernel is obtained by the identification of the state-space model. Aeroelastic analysis is conducted using state space model. The weakened wall-mounted AGARD wing 445.6 has been used for validation. An aeroelastic analysis of a NACA 65A004 composite wing model is also conducted at M=0.90 including structural nonlinearities. While highly optimized, the state-space model remains a decoupled system and has enough meaningful information about the nature of the aeroelastic system. The proposed approach is computationally efficient without losing structural/aerodynamic nonlinearities.