Design and manufacture of an aircraft wing

Abstract

A method for manufacturing an aircraft wing comprising a preliminary design phase, including iterative steps of: Determining a mesh (Mt) of the wing from descriptive geometry data (Gt), using a mesh generation and optimization tool (22); Calculating fluid flow around the wing to determine at least one parameter (Ft) expressing a target to be optimized for the wing, using a fluid mechanics solver (23); Modifying the descriptive geometry data by a reinforcement learning neural network (10) using this parameter as a reward, followed by a phase of using the descriptive geometry data for wing fabrication. Figure for the abstract: Figure 2

Images