Operator Inference-Based Model Order Reduction of Thermal Protection System Finite Element Simulations

We present an operator inference model order reduction approach for finite element simulations of reentry flight vehicle thermal protection systems. Our approach uses proper orthogonal decomposition for dimension reduction, and then infers the reduced system dynamics using a radial-basis function interpolant. A demonstration of our approach on a paraboloid geometry flying a set of trajectories starting from level flight shows that the reduced-order model is much faster then the corresponding high-fidelity full-order model, with relative errors of around 5% or less for almost all quantities of interest.