Data-Driven Modelling of Aerothermodynamic Loads during Atmospheric Re-entry

A data-driven approach based on Proper Orthogonal Decomposition is here presented and assessed with respect to the ability to make fast and accurate predictions of the aerodynamic loads acting on a spacecraft during atmospheric entry. The key contribution of the present work is to assess the ability of data-driven reduced order methods based on Principal Components Analysis in the simulation of the complex aerodynamic environment surrounding spacecraft and satellites during destructive and/or non-destructive atmospheric entry. The approach is assessed here by considering rarefied and continuum regimes for the re-entry of a 2U CubeSat test case and the ATV-cargo vehicle, respectively. A data-driven Reduced Order Model is created for the aforementioned test cases through the use of low- and high-fidelity models to inform, build and operate the resulting model. The findings of this study demonstrates the potential of Proper Orthogonal Decomposition to introduce complex flow phenomena to classical low-fidelity destructive re-entry simulations and suggests an improvement in the accuracy of the resulting re-entry kinematics with respect to high-fidelity predictions.