Internal geometric modelling of 3D woven composites: A comparison between different approaches

N. Isart, B. El Said, D. S. Ivanov, S. R. Hallett, J. A. Mayugo*, N. Blanco

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

49 Citations (Scopus)
31 Downloads (Pure)

Abstract

The mechanical behaviour of 3D woven composite materials is affected by deformations as resulting from the manufacturing process. The present study is based on comparison of three different methodologies to predict the internal yarn geometries of 3D through-thickness orthogonal interlock. The first approach idealises the geometry, which is obtained directly from manufacturing parameters assuming constant elliptical cross-sections. The second technique generates the yarn geometry from the Digital Element Method, which simulates the compaction process of the material. The last method considered is an analytical method defining longitudinal and transverse contours, which describe the undulation of fill, warp and binder yarns. The yarn geometries from the different methods are numerically analysed using voxel finite element analysis to determine the global volume fraction and the elastic properties. The results are also compared with experimental values to determine the strengths and weaknesses of each approach. The first approach is quicker than the others although the geometry is not the most accurate. The second and third method have a good match between the predicted geometry and optical micrograph of the fabric and the elastic properties are very similar for both methods.

Original languageEnglish
Pages (from-to)1219-1230
Number of pages12
JournalComposite Structures
Volume132
DOIs
Publication statusPublished - 5 Nov 2015

Keywords

  • 3D woven composites
  • Carbon fiber
  • Finite element analysis
  • Textile composites

Fingerprint Dive into the research topics of 'Internal geometric modelling of 3D woven composites: A comparison between different approaches'. Together they form a unique fingerprint.

Cite this