Speaker
Description
The models used to analyse the thermal performance of spacecraft are complex, and characterise states and behaviours of a spacecraft for all subsystems and at various levels of abstraction. Configuring thermal models is historically a technical and time-intensive process, requiring intimate knowledge of the specific model and sub-models, as well as the spacecraft, its mission and the physics underpinning heat transfer. Towards a Thermal Digital Twin (TTDT) Project develops a new approach pioneered by ESA, Airbus, the University of Leicester and Epsyl that greatly enhances the power of spacecraft thermal analysis using the digital twin concept, applied to ESA’s Solar Orbiter spacecraft. Using modern computing frameworks, our approach containerises every aspect of the thermal analysis process, creates a direct link between in-flight telemetry and the model, and radically simplifies the configuration process used to set up and execute thermal analyses. This architecture is highly scalable, making powerful stochastic methods like Monte Carlo become practical for entire models and opening a path towards routine use of Reduced Order Modelling (ROM) for thermal analysis. A more thorough understanding of the design space reduces margins and increases vehicle performance, and the reduction in engineering time needed to perform analysis translates to reduced cost and more predictable project timelines.
TTDT was an 18 month project under the ESA GSTP Programme (Invitation to Tender AO/1-10605/21/NL/KML (Activity No. 1000030753 in the “esa-star” system)) to develop an existing advanced thermal model of a complete operational spacecraft and develop it as a Thermal Digital Twin (TDT) demonstrator software environment at Technology Readiness Level (TRL) 5.
This presentation comprises a summary of the achievements, challenges and lessons learned of the project from a thermal analyst’s perspective and also includes a summary of the developments road mapped to bring the TTDT to full industrialisation at TRL 6, Product Release.
