A multiscale elasto-plastic damage model for the nonlinear behavior of 3D braided composites

Chunwang He, Jingran Ge*, Dexing Qi, Jiaying Gao, Yanfei Chen, Jun Liang, Daining Fang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

145 Citations (Scopus)

Abstract

A multiscale elasto-plastic damage model is developed to predict the nonlinear behavior of three-dimensional (3D) braided composites. In this model, the sequential multiscale method is applied to transfer the effective properties from microscale to mesoscale, and from mesoscale to macroscale. The constituents at the microscale consist of fiber, matrix and interface which are consistent with the mesoscale ones. The fiber is considered to be elastic and brittle, and the elastic damage model is applied to degrade the stiffness. For the epoxy matrix, a coupled elasto-plastic damage model is proposed to integrate the effects of plasticity and damage, and furthermore the paraboloidal yield criterion is adopted to characterize the different types of mechanical behavior in tension and compression. The bilinear constitutive relation based on the cohesive element is used to investigate the properties of interface. A user-defined material subroutine (UMAT) in the nonlinear finite element analysis software ABAQUS is written to implement the proposed model and determine the response for 3D braided composites under quasi-static tension. The numerical simulations are compared with the corresponding experiments and the results show that they agree well with each other.

Original languageEnglish
Pages (from-to)21-33
Number of pages13
JournalComposites Science and Technology
Volume171
DOIs
Publication statusPublished - 8 Feb 2019

Keywords

  • A. Polymer-matrix composites (PMCs)
  • B. Stress/strain curves
  • C. Damage mechanics
  • C. Finite element analysis (FEA)
  • C. Multiscale modeling

Fingerprint

Dive into the research topics of 'A multiscale elasto-plastic damage model for the nonlinear behavior of 3D braided composites'. Together they form a unique fingerprint.

Cite this