Data-Driven Modeling for Transonic Aeroelastic Analysis

Aeroelasticity in the transonic regime is challenging because of the strongly nonlinear phenomena involved in the formation of shock waves and flow separation. In this work, we introduce a computationally efficient framework for accurate transonic aeroelastic analysis. We use dynamic mode decomposition with control to extract surrogate models from high-fidelity computational fluid dynamics (CFD) simulations. Instead of identifying models of the full flowfield or focusing on global performance indices, we directly predict the pressure distribution on the body surface. The learned surrogate models provide information about the system's stability and can be used for control synthesis and response studies. Specific techniques are introduced to avoid spurious instabilities of the aerodynamic model. We use the high-fidelity CFD code SU2 to generate data and test our method on the benchmark supercritical wing. Our Python-based software is fully open source and will be included in the SU2 package to streamline the workflow from defining the high-fidelity aerodynamic model to creating a surrogate model for flutter analysis.