Fibre architecture optimisation in manufacturing 3D fibre reinforced composites
Researcher: Dr Xuesen Zeng
Aims
To establish a design framework for optimising 3D fibre reinforced composites by taking into account of manufacturing constraints and defects.
Progress
The optimisation framework has been implemented and evaluated.
The framework is realised through Python/MatLab scripting to interface the optimisation toolbox (MatLab), the FE pre-processor for unit cell modelling of 3D woven composites (TexGen) and the FE solver (ABAQUS).
Two optimisation algorithms have been evaluated for their performance: General Integral Genetic Algorithm and Micro Genetic Algorithm. To evaluate the performance of each algorithm with varied settings, the “experimental” study was devised by running 300 optimisation tests for each setting. Nonparametric statistical analysis of sign test and Wilcoxon signed ranks test was used to compare the two genetic algorithms. Micro Genetic Algorithm outperforms General Integral Genetic Algorithm for solving the current problem.
Two validation case studies have been formulated: (a) 3D woven optimisation for in-plane buckling resistance and (b) 3D woven optimisation design for fuselage survivability.
Evidence of Impact
This research has led to an industrial project with participants from Sigmatex, NCC and a major aerospace company.
The current research has close collaboration with the research group of Professor Prasad Potluri at University of Manchester.
