Accelerating Flight Test and Evaluation via Aerodynamic Projection-Based Reduced-Order Models

Abstract

Reduced-order models (ROMs) based on high-fidelity computational fluid dynamics (CFD) simulations have emerged as promising tools to streamline aerodynamic test and evaluation of air vehicles. This study presents the development and application of projection-based ROMs (PROMs) for the T-38 Talon aircraft, aiming to accelerate flight test campaigns while maintaining predictive accuracy. Two different CFD solvers, Stanford’s AERO-F and the Department of Defense’s Kestrel, were used to compute the aerodynamic performance of the T-38. Model outputs were validated against one another and against experimental data. The AERO-F solver was then used to construct nonlinear ROMs on a four-dimensional parameter space via the least-squares Petrov–Galerkin method augmented with energy-conserving sampling and weighting hyperreduction. These hyperreduced PROM (HPROM) predictions were compared with flight test data from a T-38C performing longitudinal maneuvers at Edwards Air Force Base, spanning a challenging range of angles of attack, Mach numbers, altitudes, and stabilator deflections. Over the approximately 90 s maneuver, the HPROMs modeled the vertical aerodynamic force acting on the aircraft with less than a 10% error relative to the flight test, correctly identified key T-38 moment and stability characteristics, and provided an 80,000-fold computational speedup over the associated CFD simulations while maintaining 96% accuracy in CFD flow variable predictions.