Efficient Nonlinear Reduced-Order Model for Computational Fluid Dynamics-Based Aeroelastic Analysis

Abstract

In this work, an efficient nonlinear state-space reduced-order model (ROM) is introduced. The nonlinear state-space ROM consists of both linear terms and nonlinear terms, the former of which is obtained based on Volterra series theory and eigensystem realization algorithm method, whereas the latter is determined by an output error-minimization procedure. Two approaches are proposed to determine the training signal for identification of the nonlinear terms. Aeroelastic analysis of a weakened wall-mounted AGARD 445.6 wing is conducted and it is demonstrated that 1) the nonlinear state-space ROM can get more accurate unsteady aeroelastic response than traditional linear Volterra-based ROM with computational fluid dynamics coupling with computational structural dynamic as a reference method for comparison; 2) the nonlinear state-space ROM performs well in range of subsonic, transonic, and supersonic; and 3) the nonlinear state-space ROM can better characterize the nonlinear aeroelastic system than traditional linear Volterra-based ROM without the need of identifying second-order Volterra kernels, which thus can be easily applied to solve engineering problems for its concise form.