This presentation was held at the 2020 NAFEMS UK Conference "Inspiring Innovation through Engineering Simulation". The conference covered topics ranging from traditional FEA and CFD, to new and emerging areas including artificial intelligence, machine learning and EDA.

Resource Abstract

Over the last decades, Automated Fibre Placement (AFP) has been gaining popularity within the aerospace, automotive and renewal energy industries and has been progressively replacing hand layup. This technology allows for higher production rates, however there are still several defects arising during the manufacturing process that prevent AFP from achieving its maximum potential. Fibre bridging at geometric features, overlapping or excessive gaps between tows, fibre misalignment, and fibre waviness are the most common defects generated during AFP deposition. In addition, wrinkle formation, temperature overshoots, under-cure and residual stress formation during the consolidation-cure process can significantly affect the quality of the manufactured part. Development of discrete composite process models has allowed predicting defect formation to some extent. However, a holistic approach integrating all individual process models in a single chain has not been developed yet. This study aims at the development and implementation of a manufacturing process model framework for the entire AFP manufacturing process chain. This will enable the investigation of the interdependencies between the different manufacturing stages, which were so far ignored, and therefore contribute towards right first-time defect-free manufacturing processes with considerable benefits in terms of time and cost. To this end, a process simulation model focusing on the cure and consolidation processes in the case of AFP was developed. The model is three dimensional and was developed using the commercial Finite Element Analysis (FEA) code Abaqus. It comprises three sub-models; a cure simulation model focusing on the heat transfer effects, a consolidation model and a residual stress model, respectively. The heat transfer model comprises the material sub-models of cure kinetics, specific heat capacity and thermal conductivity. The consolidation model is a hyper-viscoelastic model able to predict thickness variation and wrinkles formation. The residual stress model computes the development of the lamina properties as a function of the degree of cure and temperature and is able to predict residual stress formation and shape distortion. All three models were implemented using Abaqus user-defined subroutines. An L-shape laminate was manufactured in order to validate the developed model and the experimental data were compared against simulation results. It was indicated that the approach adopted here can lead to a more accurate representation of the underlying phenomena since the modelling framework is able to take into account the inputs and outputs of all the models at the same time.