Periodic Reporting for period 1 - UPBEAT (Uncertainty Prediction & Bias Elimination in Aviation Technology for Safety)
Reporting period: 2024-05-01 to 2025-10-31
Uncertainty exists throughout the life cycle of an aerostructure, from the design of individual components and material selection to manufacturing processes, operational conditions, environmental factors, and end-of-life recycling (cradle-to-cradle). To meet aviation safety and sustainability goals, Virtual Prototyping (covering design, qualification, and manufacturing) combined with Virtual Certification (VC) is emerging as a critical enabling technology.
However, variations in materials and manufacturing processes limit the reliability of purely deterministic simulations. While standards such as SAE TAHB0009A and ARP 4761 provide frameworks for aviation safety, inherent structural randomness often necessitates larger safety margins and complex certification procedures.
Understanding and quantifying these uncertainties in components and full aviation structures is essential for accurate risk and safety assessment. UPBEAT (Uncertainty Prediction & Bias Elimination in Aviation Technology for Safety) addresses this challenge by developing advanced methods for Uncertainty Quantification (UQ). These methods integrate mathematical modeling, statistics, machine learning, and simulation techniques, correlating uncertainty with severity and probability of occurrence. The ultimate goal is to enable efficient risk assessment for each failure mode of a selected engine aerostructure, the Outlet Guide Vane (OGV).
The project aims to achieve: 20-40% reduced weight and 50-70% fewer defects. Advanced predictive capabilities streamline product development, reducing qualification time by 30-40% and costs by 25-35%. In-line quality assurance (QA) support lowers manufacturing costs by 30-50% & time by 20-30%.
However, variations in materials and manufacturing processes limit the reliability of purely deterministic simulations. While standards such as SAE TAHB0009A and ARP 4761 provide frameworks for aviation safety, inherent structural randomness often necessitates larger safety margins and complex certification procedures.
Understanding and quantifying these uncertainties in components and full aviation structures is essential for accurate risk and safety assessment. UPBEAT (Uncertainty Prediction & Bias Elimination in Aviation Technology for Safety) addresses this challenge by developing advanced methods for Uncertainty Quantification (UQ). These methods integrate mathematical modeling, statistics, machine learning, and simulation techniques, correlating uncertainty with severity and probability of occurrence. The ultimate goal is to enable efficient risk assessment for each failure mode of a selected engine aerostructure, the Outlet Guide Vane (OGV).
The project aims to achieve: 20-40% reduced weight and 50-70% fewer defects. Advanced predictive capabilities streamline product development, reducing qualification time by 30-40% and costs by 25-35%. In-line quality assurance (QA) support lowers manufacturing costs by 30-50% & time by 20-30%.
Within the current duration of the UPBEAT project, two different manufacturing technologies and materials (designed to be joined in the hybrid OGV part) have been developed. A novel composite design for the vane part of the OGV has been developed with material coupon fabrication, testing and analysis well underway. Four key manufacturing parameters such as temperatures, pressures, resin flow were considered with the choice and selection for the OGV considering state-of-the-art Resin Transfer Molding (RTM) or prepreg manufacturing. Metallic Ti-6Al-4V coupons produced through laser powder bed fusion with melt-pool monitoring have been manufactured and are underway with mechanical testing. Melt pool monitoring data has been calibrated to represent build temperatures with 3D visualisations to help identify quality concerns during manufacture and aid in validating digital models. For both processes, digital models are under development considering both commercially available models (e.g. LSDyna) and inhouse developed models to capture the behaviour of the materials to loading that incorporates the variability in the materials from the manufacturing process. The next challenge is integrating these models into the uncertainty framework
The framework for modelling the propagation of uncertainty in manufacturing and final components with state-of-the-art and beyond the state-of-the-art stochastic approaches for high fidelity deterministic simulations is near completion. With activities to minimize the numerical cost of the stochastic simulations and to deal with very low levels of probability of undesired events ongoing, the next challenge is to apply these methods to processes and structural simulations of proof of concepts. Furthermore, new stochastic indexes for sensitivity analysis are being developed.
The framework for modelling the propagation of uncertainty in manufacturing and final components with state-of-the-art and beyond the state-of-the-art stochastic approaches for high fidelity deterministic simulations is near completion. With activities to minimize the numerical cost of the stochastic simulations and to deal with very low levels of probability of undesired events ongoing, the next challenge is to apply these methods to processes and structural simulations of proof of concepts. Furthermore, new stochastic indexes for sensitivity analysis are being developed.
To ensure future exploitation of UPBEAT technologies, the current SoA results will be disseminated effectively to the broader public. A conference paper has been finalized and submitted for results on the melt pool monitoring system in powder bed fusion, which will be followed by papers on modelling and simulation efforts. Furthermore, multiple papers on the composites and uncertainty propagation demonstrating novel findings (with conference attendances) are planned for the second period, where beyond the state-of-the-art stochastic approaches for high fidelity deterministic simulations are near-completion.