A Process Simulation-Based Approach for the Control of Composite Laminate Deformation During Autoclave Moulding

Abstract

In this thesis, a modelling framework for predicting material deformation during autoclave moulding
of industrial scale composite laminates is developed. The aim of the framework is to fully automate
the modelling process in an efficient way, to remove the barriers to its utilisation in an industrial
setting and enable a process simulation based iterative design methodology.
It was initially shown that a ply-by-ply modelling approach was too computationally expensive for the
requirements of the application and laminate complexity. The focus was therefore re-directed towards
an efficient homogenised material model, around which the modelling framework was constructed.
To utilise the material model, pre-processing and post-processing tools were created to automate the
simulations. A reconstruction tool, which uses the strains extracted during the homogenised
simulation to calculate the deformation of individual plies was developed, enabling an accurate
depiction of the consolidated geometry on a ply-by-ply level, in a more computationally efficient way
than the original high-fidelity approach.
A tapered thick composite laminate was designed and used to numerically study the effects of
material variability on the quality of the finished part, utilising the newly developed automation tools,
which enabled the completion of ~200 3D compaction simulations in 3 days. The results obtained
from the study were used to improve the initial laminate ply-book design, resulting in a more robust
and dimensionally compliant part. The iterative design process produced two key outcomes: it
demonstrated the use of process simulation tools in improving laminate quality by providing an
insight into the material deformation during processing; [redacted].
The initial and modified laminates were manufactured to validate
the results and demonstrate the benefits of simulation guided interventions in the early design stages.

Date of Award	9 May 2023
Original language	English
Awarding Institution	University of Bristol
Supervisor	Stephen R Hallett (Supervisor) & Jonathan P Belnoue (Supervisor)