Digital Twin for Aerodynamic and Structural Interaction

This paper details the methodology of a digital twin platform for aerodynamic load reconstruction to predict the structural response of an aeroplane wing. A laboratory model of the wing was 3D printed and tested experimentally in the T1 wind tunnel at Imperial College London. The force measurements show good agreement to the low-fidelity simulations conducted using XFOIL and JavaFoil. The main inputs to the digital twin model are; Mach number, angle of attack and altitude. The main output is the pressure field across the aerofoil surface. The structural response is calculated with the reconstructed pressure field and the main output is the strain field around the aerofoil. A Bayesian Regularization Back Propagation (BRBP) based machine learning model was found to be accurate in reconstructing the pressure distribution. Experimentally, the flow field around the aerofoil section was visualised using flow tufts. The flow field was found to be well-behaved and a small amount of separation was observed towards the trailing edge of the aerofoil model and on the tunnel floor. The structural response of the aerofoil section was calculated using the reconstructed pressure field.