Abstract
Spacecraft in Very Low Earth Orbit (VLEO) could offer low-latency communications, high-resolution Earth observation. Their proximity to Earth enables faster data transmission, improved weather and disaster tracking, and natural deorbiting for space debris compliance. However, operating at VLEO altitudes is challenging due to significant perturbations induced by Earth's spherical harmonics, atmospheric drag, solar radiation pressure, and albedo effects. While all these forces influence both orbital and attitude dynamics, atmospheric drag is the dominant factor driving orbital decay. This research develops a high-fidelity six-degree of freedom orbit and attitude propagator that build upon Gas-Surface Interaction (GSI) models to model atmospheric drag and torques. Unlike conventional approaches that use a single drag coefficient, this method leverages mesh-based satellite geometries to compute aerodynamic forces and torques, improving accuracy in capturing the complex aerodynamic environment. Additional perturbations from Earth's non-spherical gravity field, solar radiation pressure, and third-body effects from the Moon and Sun are also included. The high fidelity orbit propagator used in this work was validated using Gravity Field and Steady-State Ocean Circulation Explorer (GOCE) data with GSI parameter's described in the literature, achieving close alignment with GOCE's ephemeris during its free-fall. Building on this foundation, a fully coupled 6-DoF propagator was constructed and integrated with the Modified Chebyshev–Picard Iteration (MCPI) scheme to enable efficient solution of the coupled dynamics. A multi-resolution aerodynamic strategy was further introduced to accelerate convergence by transitioning from a low-fidelity to a high-fidelity aerodynamic model during the iterative process. The results demonstrate that MCPI can significantly reduce computational cost for coupled orbit–attitude propagation, though its performance is sensitive to initial guesses, segmentation, and node degree choices. These findings highlight both the promise of MCPI for VLEO modelling and the need for further research into segmentation and convergence strategies for MCPI used in VLEO applications.