Speaker
Description
We present a comprehensive, high-fidelity framework for long-duration thermal reliability simulations, initially developed and validated on a geostationary telecommunications satellite. Traditional methods relied on experimental formulas using minimum and maximum temperatures at specific mission phases. In contrast, our approach performs detailed thermal simulations for every single day of the satellite’s entire 17-year operational lifespan (over 6,200 cases).
Recent improvements to the radiative solver, including optimized and parallelized view factor calculations, enable the full simulation to be completed in approximately 12 days using 10 CPU cores, or 6 days with 20 cores on a standard workstation. A key technical advancement lies in the efficient parallelization of the heat transfer solution, even when daily simulations are interdependent (i.e. when the output of day d becomes the input for day d+1).
The framework accounts for time-dependent thermo-optical degradation by modeling material property evolution, allowing accurate prediction of long-term aging effects. It also incorporates transient thermal phenomena, such as localized heating due to thruster firings during station-keeping maneuvers. Additionally, realistic mission dynamics are captured through varying payload operation modes.
All these features are integrated into the modular architecture of the eTherm software, enabling seamless coupling between orbit propagation, attitude dynamics, and detailed thermal analysis. This physics-based methodology paves the way for a shift from experimental reliability assessments to simulation-driven evaluation of insurance costs, enhancing confidence in thermal margin predictions across the satellite's lifetime.
