Model-Order Reduction in Hypersonic Aerothermoelasticity Using Time-Varying Koopman Operator

Abstract

The coupled analysis between the flight dynamics, structural dynamics, heat transfer, and hypersonic aerothermodynamics, viz. AeroThermoServoElasticity (ATSE), is a key ingredient for evaluating the performance, stability, and reliability of hypersonic vehicles. A thorough performance analysis for ATSE is computationally intractable with high fidelity models for each discipline. Hence, there is a need to develop accurate reduced order models (ROM) for aerothermodynamics as well as thermoelasticity. This work exploits the latest advances in dynamical system theory to develop a reduced order model for aerothermoelastic analysis. The Time-Varying Eigensystem Realization Algorithm (TVERA) is used to identify a linear time varying (LTV) model from a high fidelity computational framework with guaranteed observability. The simulations performed show a good agreement between the reduced-order and high fidelity models. The reduced order model accurately captures the bifurcation in the aeroelastic flutter.