Real-time Aerodynamic Modeling and Control of Optimum Power-Off Glide Performance During Emergency Forced Landings
Loading...
Authors
Peterson, Jessica
Issue Date
2025
Type
Dissertation
Language
en_US
Keywords
Aerodynamic Modeling , Glide-optimization , Loss of Thrust , Power-Off Glide Performance , System Identification
Alternative Title
Abstract
This dissertation develops and validates a method to detect aircraft configuration changes, update aerodynamic performance models, and predict glide performance by creating a Kalman filter-based real-time drag model. This research utilizes theoretical modeling, system identification, and T-38 flight-test data. The framework implements a three-variable drag model, which enables real-time estimation of glide range, best-glide speed, and trajectory planning during loss-of-thrust or degraded-configuration scenarios.Loss of thrust remains one of the leading causes of mishap in general aviation and a critical hazard in single-engine military aircraft. Traditional glide-performance models rely on steady-state assumptions and precomputed data that cannot adapt to configuration changes. This research addresses those limitations by applying data-driven estimation to dynamic maneuvers, continuously updating aerodynamic coefficients and reachability predictions in real time.
The first major contribution is the extension of the classical parabolic drag model to a three-variable form: C_D=C_D0+k*C_L^2+k_L*C_L where k_L introduces a linear lift-dependent term that captures asymmetry in the drag polar when minimum drag occurs at nonzero lift.
The second contribution derives analytical expressions for optimum glide performance, yielding an updated maximum-glide efficiency relation: (L/D)_max=1/(2√(kC_D0 )+k_L ).
The third contribution demonstrates that configuration changes affect all three coefficients (C_D0, k, and k_L), not merely the traditionally accepted change to parasite drag only.
The fourth develops a real-time Kalman filter estimation framework that continuously updates aerodynamic parameters using adaptive covariance weighting and configuration-change detection logic.
The fifth applies the real-time model to glide-optimization and path-planning algorithms developed in collaboration with the University of North Carolina at Charlotte.
Validated with T-38 flight-test data collected at the U.S. Air Force Test Pilot School, this research bridges steady-state aerodynamic characterization and dynamic, in-flight adaptation, advancing real-time safety, decision support, and autonomous guidance during loss-of-thrust emergencies.