Instant CFD demonstrator on NACA0012 design and ONERA M6 wing cases (publicly available)

Demonstrators of (i) a hybrid deep learning and multi-ridge framework that can predict entire RANS flow-field instanteously (ii) dimension reduction methods for flow-fields rooted in computational optimal transport and manifold optimisation; and (iii) lambda-DNN based approach for fluid flow-field and interactions. The outputs will directly feed into the robust design of a high bypass-ratio fan in WP4/WP5, along with the Digital Twin of the heat exchanger in task 6.3. The lambda-DNN approach will be deployed in WP3/4. Linked to methodologies devloped in task 2.1.