The prediction of wrinkle formation in non-crimp fabrics during double diaphragm forming

Yu, Fei (2022) The prediction of wrinkle formation in non-crimp fabrics during double diaphragm forming. PhD thesis, University of Nottingham.

PDF (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Available under Licence Creative Commons Attribution.
Download (24MB) | Preview


Liquid composite moulding (LCM) processes are an economical alternative to autoclave-cured prepreg, as intermediate material forms are less expensive, capital equipment costs are lower and cycle times are shorter. LCM is suitable for producing aerospace quality components, but the process chain is dependent on a separate preforming process to convert 2D fibre formats into complex 3D shapes prior to moulding. Preforming is difficult to automate to ensure defect-free architectures, as multiple plies are often formed at the same time according to predefined forming loads and constraints. Corrections to the draping direction, force and sequence can be easily refined during manual hand layup, as the laminator works on one ply at a time. Corrections to the forming sequence cannot be easily made during automated forming, therefore numerical models are required during the design phase to refine process parameters to ensure that defects do not evolve.

This thesis seeks to develop a robust simulation methodology for modelling the forming behaviour of biaxial fabrics, in order to enable effective identification of forming related defects and assist in process design for improving the quality of preforms. Research has been conducted in three main areas:

(I) Material characterisation and model development.

A macroscale constitutive model has been developed to simulate the forming behaviour of biaxial fabrics, based on an explicit finite element scheme, incorporating the effects of bending stiffness to predict wrinkling. Cantilever tests were employed to characterise the bending behaviours of a twill-weave woven fabric and a non-crimp fabric (NCF) with pillar stitches, providing linear (a constant rigidity) and nonlinear bending stiffness models to represent the fabric materials. Experimental and numerical studies have shown that the bending behaviour of the fabrics is nonlinear, which is dependent on the fibre curvature along the bending direction. Forming simulations using a constant bending stiffness from the standard cantilever test (BS EN ISO 9073-7; 1998) produced unrealistic predictions for fabric bending and wrinkling behaviours, while the nonlinear model produced more accurate forming induced wrinkle patterns compared to the experimental data. The nodal distance between the deformed fabric mesh and the tool surface was identified to be a suitable method to locate areas containing out-of-plane defects, using the principal curvature to further isolate wrinkles from areas of fabric bridging (poor conformity).

(II) Multi-resolution modelling approach for defect identification.

A multi-resolution modelling strategy was developed for determining forming induced defects in large-scale DDF components, using a macroscopic global-to-local sub-modelling technique. The fabric constitutive model developed in Stage (I) above was used for predicting macro-scale defects (i.e. wrinkling and bridging defects at the ply level) during double diaphragm forming (DDF) of a generic geometry comprising local changes in cross-sectional shape. Comparisons between simulations and experimental results confirmed the accuracy of the forming model, but the runtime of the full-scale shell-element model was found to be impractical. Therefore, a multi-resolution modelling strategy was developed to improve the overall computational efficiency. Areas containing potential defects were initially determined by a full-scale global simulation using a coarse membrane-element mesh (element edge length of ~5mm). Results with higher resolution (wrinkle amplitude of ~1mm) and more realistic shapes were subsequently obtained by local sub modelling, using higher order shell-element meshes, based on boundary conditions derived from the global simulation.

The applied methodology enabled an 87% saving in the runtime compared to the high fidelity full-scale shell-element model for the same geometry, with the length of wrinkles and area of fabric bridging predicted within 10% compared to experimental data. These reductions will become more significant as the overall length scale is increased for components produced by DDF, as defects will tend to be within more localised regions.

(III) Defect formation and mitigation during multi-ply NCF forming.

Forming experiments and simulations were performed to investigate the mechanism of wrinkling in multi-layered NCF plies during DDF. Simulation results indicate that in-plane fibre compression, caused by dissimilar shear deformation between adjacent plies, can lead to out-of-plane wrinkles, where the wrinkle length is a function of the relative fibre angle at the ply-ply contact interface. The most severe wrinkles occurred when the inter-ply angle was 45° for a multi-ply biaxial NCF preform. Numerical and experimental studies have shown that out-of-plane wrinkles are sensitive to the friction resistance between NCF plies and therefore lubricating the fibres can minimise wrinkling defects caused by dissimilar inter-ply deformation.

In summary, results from the first part of the thesis (Chapter 3 and Chapter 4) demonstrate the importance of incorporating a curvature-dependent bending behaviour into fabric constitutive modelling for correctly predicting forming behaviour of bi-axial fabrics. The multi-resolution forming simulation strategy (Chapter 5) extends the capability of the proposed fabric model to predict macroscale defects in large structures more efficiently. The study on multi-ply fabric forming (Chapter 6) provides further understanding about the effect of inter-ply sliding on ply wrinkling, enabling a feasible solution for wrinkle mitigation. The results from this work can be directly extended for industrial application to improve the performance of composite structures made via fabric preforms.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Harper, Lee
Warrior, Nicholas
Chen, Shuai
Keywords: Fabrics/Textiles; Industrial fabrics; Forming; Finite element analysis (FEA); Material characterisation; Carbon fiber-reinforced plastics; Manufactures,Defects
Subjects: T Technology > TS Manufactures
Faculties/Schools: UK Campuses > Faculty of Engineering
Item ID: 68274
Depositing User: Yu, Fei
Date Deposited: 15 Mar 2022 04:40
Last Modified: 15 Mar 2022 04:40

Actions (Archive Staff Only)

Edit View Edit View