The aerospace engineering field is increasingly focusing on sustainability, influencing the entire aircraft lifecycle from design to operation. Central to this shift are virtual prototypes and digital twins, which provide dynamic, time-based digital representations of aircraft systems, enabling virtual qualification, real-time monitoring, predictive maintenance, and informed decision-making to enhance efficiency and safety.

Data-driven surrogate models (DDSMs) play a crucial role in developing these digital twins by leveraging machine learning (ML) techniques to approximate system behaviors based on operational data or high-fidelity simulations. DDSMs significantly reduce computational demands but also introduce uncertainties due to variations in ML predictions and data quality. Addressing these uncertainties is vital for creating reliable and high-quality digital twins.

QUASAR's aim is to develop and identify robust methods for quantifying uncertainties in complex physical systems that rely on interconnected DDSMs, combined with analytical and physical models. The project is funded for two years by Baden-Württemberg’s program "Luft- und Raumfahrttechnik 2050, Nachhaltig-Digital-Kooperativ".

The research uses a gas turbine-powered model helicopter as a case study for dynamically operated aircraft, requiring collaboration across multiple disciplines, including engine performance modeling, fluid dynamics, thermodynamics, system reliability, and ML assurance. Experts from several institutes at the University of Stuttgart  - ILS, IAG, ITRL and ILA  - are collaborating to address multilevel uncertainties, spanning from high-fidelity computer modeling and simulations (LES) to data-driven lower fidelity models (RANS), and culminating in a comprehensive digital twin model. This digital twin integrates data-driven surrogate models and analytical approaches to capture the complexities of system-of-systems behavior. This interdisciplinary effort ensures that QUASAR can effectively address the challenges of modeling complex aerospace systems.