« HOME
Queen's University Belfast Logo
Montage
Centre of Excellence of Integrated Aircraft Technology Logo
University Sites 
- CEIAT 
- RESEARCH ACTIVITY 
- Multi-Disciplinary Design Optimisation

-
Postgraduate Research
Menu Image
Automated Design
Menu Image
CRESENDO Framework 7
Menu Image
Impact Modelling in Complex Systems
Menu Image
Multi-Disciplinary Design Optimisation
Menu Image
Shape Sensitivity Analysis and Parametric CAD Models
Menu Image
SILOET Project
Menu Image
Sustainable Transport
Menu Image
Systems Engineering
Menu Image
Technology Infusion
Menu Image
Virtual Test Systems
Menu Image
Visual Analytics
Menu Image
Systems Integration and Cost Modelling
Menu Image
Digital Manufacturing
Menu Image
Generator Aeroacoustics
Menu Image
Aerospace Automation
Menu Image
Multi-Disciplinary Design Optimisation

Multi-Disciplinary Design Optimisation

Researchers

Dr. Theresa Robinson

Mr. David Lisk

Project Summary

Multi-disciplinary design optimisation uses optimisation methods in order to solve design problems.  The problems are converted into math problems and then solved computationally.   As there are typically many design variables and several objectives to be optimised, it enables designers to incorporate all of the relevant disciplines simultaneously. However this increases the complexity of the problem.  As the disciplinary models are often very complex a significant amount of time can be required to obtain a single evaluation.  The result of multi-disciplinary design optimisation is a valuable range of options that can be considered for use in the design process. 

Publications

 Paper: Optimisation of Supersonic Projectiles using Adaptive Sampling of the Pareto Front

David Lisk, PhD Student, School of Mechanical and Aerospace Engineering

Theresa Robinson, Lecturer, School of Mechanical and Aerospace Engineering

Des Robinson, Senior Lecturer, School of Planning, Architecture and Civil Engineering

Abstract

This paper describes the implementation and demonstrates the application of an optimisation framework for supersonic projectiles.  A surrogate model is constructed using Kriging from initial sample data produced by two aero-predictive codes (CART3D and MISL3).  A multi-objective evolutionary algorithm (NSGA-II) was used to determine the Pareto front.  By applying an adaptive sampling algorithm, the surrogate model was refined in the region of optimal designs.  Analysis of the results showed that the point of maximum lateral acceleration intersects all of the two-dimensional Pareto fronts.  This design consists of a small tail and large canard with maximum deflection.  Although the time-to-target and roll moment variation objectives would appear unrelated, the analysis showed that they both drive the design in the same direction; towards a small tail and small canard aileron deflection.  Static stability was the main driver for an increase in tail size.  Overall, there were no significant discontinuities in the Pareto fronts and it was established that for this configuration of projectile, below 10g lateral acceleration there was little possible improvement to any of the other objectives.