Speaker
Description
Thermal uncertainty margin philosophy has been one of the critical factors in ensuring space mission success since the dawn of space era. Building on a review of its historical evolution, this work aims to quantitatively assess the current state of thermal margins, by comparing in-flight temperatures with thermal model predictions for the Sentinel 1C satellite.
Specifically, the platform’s detailed thermal mathematical model (TMM) has been updated with registered telemetry data – including unit dissipations, attitude profile and heater line statuses – collected during Summer Solstice period. The refined model’s predictions were benchmarked against flight thermistor measurements from the same interval. The same process was then repeated on the non – correlated thermal mathematical model of S1C, isolating the impact of the correlation exercise.
Statistical post-processing of the results – focusing on bias, standard deviation, Uncertainty Flight Prediction (UFP) and its associated confidence level – revealed discrepancies between predicted and in-flight temperatures on both the mathematical models, but highlighted the critical role of the correlation exercise in reducing satellite UFP. In particular, considering a confidence level of 98% the calculated uncertainty decreases of 46% (from 13°C to 7 °C).
These findings highlighted persistent gaps in predictive accuracy during on-orbit operations, primarily driven by the inability of accurately replicate the unit dissipations (due to missing dedicated telemetry) and missing harness thermal effect. Conversely, the results outlined the vital role of TMM correlation in narrowing the uncertainty on temperature predictions.