Decomposed Reduced-Order Modeling for Flutter and LCO Predictions in Highly Nonlinear Systems

Modeling aircraft aeroelastic response is an incredibly challenging process fraught with many questions regarding the approaches and assumptions in both structural and aerodynamic analyses.nbsp; Modeling 3-D, full scale, fully coupled, aerodynamic and structural responses with high-fidelity computational approaches is only viable for evaluating a few conditions within the flight envelope, but intractable for defining a flutter boundary over a range of flight speeds.nbsp; To alleviate these challenges researchers have developed reduced order models to facilitate the aerodynamic calculations at a fraction of the cost.nbsp; These methods are incredibly powerful, but present significant problems when attempting to apply in the case of highly nonlinear flows as they can be costly to maintain fidelity required in the modeled response.nbsp; These nonlinear flows play a critical role in the aeroelastic response and as such require that the reduced order models provide a high level of fidelity.nbsp;The proposed research will demonstrate a framework that decomposes nonlinear aerodynamic responses, in the form of Generalized Aerodynamic Forces based upon dynamical models which are then extended to the nonlinear range based upon the concept of a Volterra series.nbsp; By decomposing the Volterra series into the linear and nonlinear parts a significant cost savings can be leveraged as the linear terms can remain fixed for a given choice of flow parameters and the Volterra series need only serve to reproduce the nonlinear response of the system.nbsp; By taking this approach to modeling the aerodynamics we hypothesize an improvement in the reduced order modelrsquo;s ability to reproduce an accurate nonlinear representation of the aerodynamics at a greatly reduced cost.nbsp; This will aid in both increasing the fidelity of aeroelastic predictions and provide a valuable resource to utilize in the development of aeroservoelastic control logic.