Multi-fidelity Design with Optimization Guided Incremental Decisions

Abstract

The aerospace industry is continually seeking novel, sustainable, and efficient aircraft designs to address global emissions reduction goals and growing air travel demands. In this context, innovative configurations are being explored to provide sustainable air transportation in diverse markets. However, traditional design processes based on empirical methods and expert knowledge may be inadequate for these novel configurations. This thesis introduces a design strategy that utilizes physics-based models and Multi-Disciplinary Analysis and Optimization (MDAO) as a framework to approach designs without precedents. The proposed strategy is demonstrated through a design example of an electric regional passenger aircraft, where both analysis fidelity and geometry representation are incrementally refined. The design example culminates in an optimization formulation that determines the optimal wing, fuselage, and tail surface geometries, as well as the ideal flight conditions for takeoff, climb, cruise, and landing. The final formulation incorporates two high-fidelity models within the optimization loop: MSES for airfoil analysis at multiple spanwise locations, integrated over the wing using a lifting line method, and a finite element structural analysis for sizing optimal wing structural components.