Multiscale modelling strategy for a novel wingbox structure with increased bend-twist coupling
File(s)
Published version
Author(s)
Miranda, Mário
Li, Runze
Iorga, Lucian
Pinho, Silvestre T
Type
Journal Article
Abstract
This paper presents a novel multiscale modelling strategy for the early stage design of novel composite components. It includes the use of a global-local simulation approach followed by training a physics guided artificial neural network to predict composite failure indices from global-level loads. As a proof of concept, a non-conventional wingbox using the Z-Beam concept, capable of providing tailored bend-twist coupling for aeroelastic tailoring is analysed. The process begins by the application of strain-based global loads on a fully parametrised Z-Beam wingbox, after which, a structural detail of interest is analysed using a global-local approach and an evolved LaRC05 composite failure criterion. A database is created based on the repetition of the previous steps and is used to train a physics guided artificial neural network. At the end, it is observed that the proposed methodology is able to accurately predict composite failure based on global loads and design variables in a swift manner, proving itself as a valuable asset for the early design stages of novel composite concepts.
Date Issued
2025-10-15
Date Acceptance
2025-05-15
Citation
Composite Structures, 2025, 370
ISSN
0263-8223
Publisher
Elsevier
Journal / Book Title
Composite Structures
Volume
370
Copyright Statement
© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
License URL
Publication Status
Published
Article Number
119291
Date Publish Online
2025-06-01