Developing cutting edge digital twin technologies for the aerospace sector by synthesizing data with physics-based models, facilitating improved aerothermal inference for asset performance, design and manufacturing.
Explaining the science
This work is a multi-disciplinary marriage of computer science, computational statistics and aerothermodynamics, and is guided by the need for real industrial impact. The project brings together computer aided design (CAD) models, real-time sensor data and low-fidelity aerodynamic models for extracting value. In this respect, the work builds on the data-centric methodological advances pursued within the related 'Digital twins in aeronautics' project.
The initial exploration of the feasibility of using such system for aeronautical design has been investigated through development of the AeroVR system where the researchers presented a Virtual Reality (VR) aerospace design environment with the objective of aiding the component aerodynamic design process by interactively visualising performance and geometry. This virtual environment uses ideas from parameter-space dimension reduction to enhance the exploration and exploitation of the design space (see Tadeja et al.).
The scope of the proposed research would include further enhancement and expansion developed on top of the existing system. The research will consider the problem of comparing the health of aeroengines across a fleet of an airline. The aim is to develop an immersive environment where VR can be used to compare the performance of different engines, ascertain the cause of variability in performance, and make recommendations on component overhaul and repairs.
The virtual environment would use 3D geometric computer-aided design (CAD) models of the engine, paired with a computational representation of the flow through the engine. This computational representation would be based on sensor data that is acquired from a range of temperature and pressure sensors installed in each engine. Gaussian process regression will be used to build engine models and to aid in anomaly detection and detailed component-wise monitoring. This would facilitate real-time, dynamic insight into aeroengine prognosis and diagnosis.
The main goal of this project is to research, test and develop a VR-based tool for immersive, interactive visualization for aerospace design and digital twinning. The immersive environment will be used to compare the performance of different engines, ascertain the cause of variability in performance, and make recommendations on component overhaul and repairs.
As mentioned above, the current focus is on use cases within the aeroengine and aircraft industries.