Skip to main navigation Skip to search Skip to main content

Bending performance and failure mechanism prediction of 3D angle-interlock woven composites via full-scale finite element modeling

  • Yue ying Yin
  • , Dian sen Li*
  • , Yi fan Zhang
  • , Li Chen
  • , Lei Jiang
  • , Stepan V. Lomov
  • , Frederik Desplentere
  • *Corresponding author for this work
  • Beihang University
  • KU Leuven
  • Tiangong University

Research output: Contribution to journalArticlepeer-review

Abstract

Parametric full-scale finite element models were established based on actual yarn structure to predict the three-point bending properties and progressive failure process of 3D angle-interlock woven composites(3DAWCs) under different yarn densities. The simulation results are in good agreement with experimental data, verifying the validity of models. The results show that yarn density has a strong correlation with the bending behavior of 3DAWCs. The bending performance is positively correlated with warp yarn density, but first positive then negative with weft density. During the bending process, the warp yarns are the main load-bearing components. Damage first occurs on the outer warp yarns, and propagates towards the interior of material along the thickness direction. The weft yarns and the matrix bear part of the load, and and failed due to shear and compressive effects respectively. Appropriately increasing the yarn density is beneficial for making the stress distribution more uniform, while an excessively high weft yarn density will lead to stress concentration, thus causing a decline in the bending performance of 3DAWCs.

Original languageEnglish
Article number120343
JournalComposite Structures
Volume387
DOIs
StatePublished - May 2026

Keywords

  • 3D angle-interlock woven composites
  • Bending performance
  • Failure mechanism
  • Finite element analysis

Fingerprint

Dive into the research topics of 'Bending performance and failure mechanism prediction of 3D angle-interlock woven composites via full-scale finite element modeling'. Together they form a unique fingerprint.

Cite this