Computationally efficient process modelling of automated fibre placement

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

The use of composites has become increasingly important in reducing emissions and increasing sustainability. However, the current manufacturing landscape is dominated by manual layup, which offers limited scalability for the ever increasing demand for composite parts. Automated Fibre Placement (AFP) is a promising process for laying up unconsolidated slit-tapes onto a tool surface using heat and compaction, especially for complex geometries like engine fan blades. However, further research is necessary to optimise AFP technology for greater accuracy, speed, and scalability, to meet the increasing demand for composite parts and achieve sustainability goals by reducing waste during manufacture and emissions during service. One key area of research is predicting the formation and evolution of variabilities and part conformance during by AFP, which is difficult due to the complex behaviour of the tape and dynamic interactions during the process. This thesis aims to develop physics-based simulation platforms to investigate material-process interaction during AFP and enable exploration of the entire processing window for assessing part quality and conformity, concentrating on controlling preform thickness and eliminating steering induced defects. The ultimate goal is to virtually derive a process window, identify areas for further process optimisation, and validate the results through small-scale manufacturing trials, as demonstrated in this work. Current numerical solutions for AFP either lack proper material definition or require extensive computation power, making them unsuitable for design.


The first three chapters explore the underlying motivations driving the adoption of automated manufacturing processes, specifically AFP, within the realm of composite manufacturing, and its pivotal role in a sustainable future.
The work then delves into the imperative need for process simulations, offering an extensive review of relevant literature while also highlighting the existing limitations associated with current simulation platforms. Chapter four addresses some of these said limitations, with a key focus on improving computational efficiency of the simulation platform. Chapters four and five delve into the practical applications of this improved platform. Chapter four centres on bulk preform compaction, while Chapter five focuses on the quality of steered tow layup across a wide range of processing conditions. Finally, Chapter six provides a comprehensive summary of the thesis, highlighting key findings and paving the way for future research endeavours.


The AFP process simulation developed in this thesis represents a significant advancement, incorporating a wide range of temperature and rate effects with much less computational effort. The novelty of proposed numerical scheme lies in the computationally efficient strategies adopted along with complete parameterisation of the process and machine, making it a rapid and versatile design tool. This thesis provides a blueprint on how to improve finite element simulations and make process models more viable as design tools.
Date of Award23 Jan 2024
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorJonathan P Belnoue (Supervisor)

Cite this

'