Data-Driven Multidisciplinary Design and Optimisation of Future Aircraft

This thesis addresses the challenge of aerodynamic design optimisation and multi-disciplinary optimisation by introducing novel approaches that leverage model-order reduction (MOR) techniques and advanced machine learning (ML) methodologies to significantly enhance computational efficiency compared to traditional Computational Fluid Dynamics (CFD). The core innovation lies in developing a non-intrusive machine learning framework for building reduced order models (ROMs) using neural networks. This facilitates the rapid and accurate evaluation of aerodynamic characteristics, which is crucial for optimizing aircraft designs. The methods proposed accurately model subsonic and transonic aerodynamics, as demonstrated by detailed studies across various operating scenarios. It achieves accuracy comparable to full-order modelling (FOM) while significantly reducing computational cost. Significant improvements are shown in predicting aerodynamic forces and shock locations with reduced computational effort. By enhancing efficiency without compromising accuracy, this work represents a leap forward in aerodynamic modelling. These methods benefit the aerospace industry by accelerating design optimisation, enabling rapid aerodynamic insights for improved decision-making, and supporting cost-effective analysis and exploration of complex designs. This research not only advances computational techniques but also provides practical tools to address intricate aerodynamic challenges, ultimately contributing to more innovative and effective aircraft design. These advantages align with the objectives of the aerospace industry to develop more efficient, sustainable, and advanced aircraft solutions.

Name dc:type.qualificationname

PhD

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

De Montfort University

Year dc:date.issued

2025

Author dc:creator

• Moni, Abhijith

Licence

• https://dora.dmu.ac.uk/bitstreams/103b2d47-f164-46da-9f58-4380843a1b39/download