Very Low Earth Orbits (VLEO) offer unique opportunities for satellite missions but remain a challenging regime due to variable environmental conditions and interaction with the residual atmosphere, which drastically limits orbital lifetimes. In this environment, aerodynamic lift can be exploited for orbit control, enabling manoeuvres such as inclination changes that would otherwise require propellant. However, the effective control authority of lift is highly uncertain, as it strongly depends on gas-surface interaction. Simplified scattering models risk misrepresenting both achievable lift and the associated drag penalties. This work investigates the extent to which inclination changes can be achieved solely through aerodynamic control using a machine-learning-based GSI model derived from molecular dynamics simulations, recently developed and published by some of the authors. Integrated within an optimization framework, this model demonstrates that lift-driven and decay-optimal inclination adaptations are feasible, though modest in scale. The results highlight promising directions for future satellite design optimization, with the new GSI model capturing beneficial aerodynamic characteristics. Overall, the findings establish aerodynamic lift as a limited yet practical control mechanism in VLEO, with drag as a cost that can be actively optimized in the manoeuvre design. • Decay-optimal aerodynamic manoeuvres change inclination with minimum altitude loss. • Machine-learning gas-surface interaction model integrated into ADBSat. • Molecular scattering physics considered in satellite-level aerodynamics. • Revealing favourable aerodynamic properties for future satellite designs. • Manoeuvring envelope determined for exemplary satellite.
Turco et al. (Sun,) studied this question.