Multiple Level Reduced Order Modeling for Ship Airwake Modeling and Simulation

The overall technical objective of the Phase I effort is to develop a Multi-Level Reduced Order Model (ML-ROM) for the airwake flow field to reduce the storage requirements by a factor of 1000 or more. This overall reduction will be achieved cumulatively by reducing the data over the spatial dimension (Level 1), over the space of the parameter (wind velocity, wind-over-deck angle) values (Level 2), and over the temporal dimension (Level 3). The Proper Orthogonal Decomposition (POD) method will be employed as the Level 1 ROM approach to reduce the data while optimally retaining the spatial resolution of the airwake flow field velocities. Level 2 reduction will next be carried out over the set of these parameters through the construction of diffusion-maps. Level 3 reduction will be carried out through an autoregressive modeling extracting the temporal variations in the frequency band of interest for aircraft/helicopter flight dynamics simulation.This ML-ROM will be implemented into one of ZONA Technologys existing software system, re-designated as zPOD+ and will be validated on in-house Computational Fluid Dynamics (CFD) data and Government Furnished Information (GFI) CFD airwake datasets.