High-precision physics-based radiation force models for the Galileo spacecraft

We present two new high-precision physics-based radiation force models for the In-Orbit Validation (IOV) and Full Operational Capability (FOC) spacecraft (s/c) of the Galileo Global Navigation Satellite System (GNSS). In both cases, the s/c bus surfaces are cov-ered in material types, i.e., Laser Retro-reflector Array (LRA), Optical Surface Reflector (OSR) and Single-Layer Insulation (SLI) cov-erings, that were either not encountered or not specifically dealt with in earlier work. To address this, a number of modelling enhancements were proposed and tested, including: a specific model to account for the direct and reflected solar radiation force for LRA surfaces; a design update of the bus model computation process to allow for more than one insulation material; a specific thermal force model for OSR surfaces; a thermal force model for the Navigation Antenna (NAVANT) surface that includes a temperature model derived from on-orbit temperature measurements; and force models to account for thermal emissions from radiator panels on the thorn +X and +/- Y surfaces for both IOV and FOC, and on the -Z surface for FOC only. In the UCL2+ model each of these effects are accounted for. The theoretical impact of each modelling concept introduced is assessed, individually, by considering the magnitude of its effect in acceleration-space. The impact on orbit accuracy is confirmed through a rigorous set of Precise Orbit Determination (POD) validation tests, in which observations from all active Galileo s/c over two full years, 2017 and 2018, including during eclipsing periods, are included in the analysis. The UCL2+ approach results in day boundary discontinuities of 22 mm, 17 mm and 27 mm in the radial, across-track and along-track components, respectively. Analysis of the one-way Satellite Laser Ranging (SLR) residuals suggests that radial accuracy at better than 1 cm (3.7 mm mean residuals) and precision at better than 2 cm (17 mm root mean square (rms) error) is achievable with the UCL2+ model. (C) 2022 COSPAR. Published by Elsevier B.V.