A DoE-based approach for the implementation of structural surrogate models in the early stage design of box-wing aircraft

TL;DR

This paper introduces a Design of Experiments (DoE)-based method to develop surrogate models for early-stage structural mass estimation of innovative box-wing aircraft, addressing limitations of existing approaches.

Contribution

It proposes a novel DoE-based framework for creating surrogate models tailored to the structural analysis of disruptive aircraft configurations like box-wings.

Findings

Validated surrogate models for wing sizing

Effective optimization framework integration

Successful application to regional aircraft case

Abstract

One of the possible ways to face the challenge of reducing the environmental impact of aviation, without limiting the growth of air transport, is the introduction of more efficient, radically different aircraft architectures. Among these, the box-wing one represents a promising solution, at least in the case of its application to short-to-medium haul aircraft, which, according to the achievement of the H2020 project "PARSIFAL", would bring to a 20% reduction in terms of emitted CO2 per passenger-kilometre. The present paper faces the problem of estimating the structural mass of such a disruptive configuration in the early stages of the design, underlining the limitations in this capability of the approaches available by literature and proposing a DoE-based approach to define surrogate models suitable for such purpose. A test case from the project "PARSIFAL" is used for the first conception of the approach, starting from the Finite Element Model parametrization, then followed by the construction of a database of FEM results, hence introducing the regression models and implementing them in an optimization framework. Results achieved are investigated in order to validate both the wing sizing and the optimization procedure. Finally, an additional test case resulting from the application of the box-wing layout to the regional aircraft category within the Italian research project "PROSIB", is briefly presented to further assess the capabilities of the proposed approach.