Periodic Neural Mapping for Unsteady Rotor-Blade Pressure and Aeroelastic Load Prediction

This work introduces Periodic Fourier Neural Mapping (p-FNM), a neural-operator framework designed for turbomachinery applications to predict unsteady pressure distributions on turbine rotor blades. The method leverages periodic Fourier features to efficiently model the temporal evolution of the pressure field without requiring expensive high-fidelity CFD simulations. It targets aeroelastic Quantities of Interest (QoI) that depend sensitively on pressure field dynamics, providing a computationally efficient surrogate for turbomachinery design workflows.