Toward Real-Time Aero-Icing Simulation of Complete Aircraft via FENSAP-ICE

Three-dimensional fully viscous turbulent aero-icing flow simulation remains too computationally intensive when broad parametric studies are needed, such as during a certification process. In addition, the introduction of realistic icing effects for training pilots in simulators clearly lags behind in terms of taking advantage of computational fluid dynamics. To make such simulations more practical, this work presents a reduced-order modeling, based on the proper orthogonal decomposition method, that predicts a wide swath of approximate flowfields and ice shapes based on a limited number of obtained from high-fidelity computations. Modes are extracted from these snapshots and used to reconstruct the computational fluid dynamics field, and/or the aerodynamic coefficients, and/ or the ice shapes for other conditions within the range. This reduces calculation times by two to three orders of magnitude from the full three-dimensional ones, enabling a more complete map of the performance of an iced aircraft over a wide range of flight and weather conditions to be used in its certification and pilot training.