A unified multi-fidelity aero-structural design framework for novel aircraft configurations

Abstract

This study presents a novel framework that integrates multi-disciplinary simulation data, robust multi-objective optimization (MOO), and interactive visual analytics to balance multiple objectives and cope with design uncertainties. Using the CRM wing and ATI narrowbody aircraft geometry as a test case, aerodynamic (e.g., lift-to-drag ratio) and structural (e.g., wing mass) performance measures are combined within a Kriging-based surrogate modelling approach. Monte Carlo Simulation is employed to propagate design uncertainties, while NSGA-III guides the exploration of Pareto-optimal solutions using a robust optimisation framework. An HTML-based dashboard, powered by Python servers, allows users to dynamically inspect trade-offs, filter design alternatives, and compare competing objectives in near real-time. Results indicate that increasing the wing aspect ratio and reducing skin root thickness generally enhances aerodynamic efficiency and flight range, although structural weight and associated uncertainties also rise. By visualizing such trade-offs interactively, designers gain insight into where higher fidelity analyses or additional data collection may be most beneficial, and which configurations deliver robust performance under probable variations. The resulting tool fosters a more efficient and informed engineering design process by consolidating data, computations, and insights within a single, accessible environment.