Data-Physics Dual-Driven (Hybrid PINN) Modelling for Next-Generation Aerospace Structural Dynamics
About the Project
Are you passionate about the fusion of Artificial Intelligence and structural dynamics? Do you want to contribute to cutting-edge research that will transform aircraft structural health monitoring and real-time digital twins?
Applications are invited for a 3.5 to 4-year PhD position focusing on "Data-Physics Dual-Driven Modelling for Next-Generation Aerospace Structural Dynamics" in the research group of Dr Jun Wu at the University of Birmingham.
Conventional structural dynamics relies heavily on high-fidelity Finite Element Analysis (FEA), which is computationally prohibitive for real-time applications. On the other hand, pure data-driven AI models lack physical constraints and generalizability. This PhD project aims to bridge this gap by developing a novel Hybrid Physics-Informed Neural Network (Hybrid PINN) framework. The framework will integrate sparse, noisy experimental sensor data with governing partial differential equations (PDEs) of structural dynamics to achieve real-time, full-field dynamic response reconstruction and material parameter identification for complex aircraft components.
Project details:
The successful candidate will adopt an integrated computational–experimental approach, spanning from theoretical algorithm development to laboratory validation. The research will systematically progress through four clear phases, scaling up in structural complexity:
Phase 1: Single Structural Component
You will start with foundational theory, training the PINN to solve classic vibration equations for a single isolated component (e.g., a beam or a plate) to accurately map its dynamic behaviours.
Phase 2: Multi-Structure Assembly & Coupling
You will advance the framework to handle multi-structure coupling. You will develop "interface boundaries" in the neural network to seamlessly connect different components. The network must learn to balance force and displacement transmissions across these coupled junctions.
Phase 3: Geometry Complexity & Real-World Non-linearities
You will scale the model to realistic aircraft assemblies featuring geometric complexities and physical non-linearities.
Phase 4: Laboratory Validation & Real-Time Deployment
You will conduct hands-on laboratory testing on a physical, multi-component assembled wing prototype. This noisy, real-world experimental data will be used to calibrate your coupled PINN model, which will finally be deployed onto edge-computing hardware for millisecond-level digital twin monitoring.
This project directly aligns with the future of digital aviation, smart structures, and digital twin technology. You will join a vibrant, multidisciplinary research team with access to state-of-the-art computational clusters and dedicated experimental dynamics facilities.
Find Your Best Opportunity
Tell them AcademicJobs.com sent you!
