Multiscale image-based modeling of progressive damage in 3D5D braided composites with yarn-reduction

Xiaodong Liu, Jingran Ge*, Xiaoxu Wang, Kun Qian, Jun Liang, Diantang Zhang

*此作品的通讯作者

科研成果: 期刊稿件文章同行评审

5 引用 (Scopus)

摘要

The special-shape fiber-reinforced composites are improbable to be represented by the representative volume element (RVE). Hence, a multiscale image-based model is proposed to reveal the progressive damage in 3D5D braided composites with/without yarn-reduction. The microscale and mesoscale models of the materials are reconstructed by Micro-CT, considering the yarns’ twists, global compression, and yarns’ misalignment. Subsequently, the continuous damage method and the cohesive-zone model are proposed to characterize the damage behavior of the microscale and mesoscale constituents. These constitutive models are implemented by a user-defined subroutine UMAT in ABAQUS. Finally, based on the homogenization procedure and the multiscale analysis method, the effects of structures on microscale, mesoscale, and macroscale properties of 3D5D braided composites without/with yarn-reduction are analyzed sequentially. The results show that cross-section variations and twists in the yarns modify the damage initiation. The wedge-shaped stress migration phenomenon occurs near the yarn-reduction point. Besides, the macroscopic simulation of strain fields and the damage morphology agrees well with the DIC measurements and Micro-CT results. The established multiscale framework is expected to predict the progressive damage of 3D5D braided composites with/without yarn-reduction and provide the theoretical basis for the structure design.

源语言英语
文章编号112236
期刊International Journal of Solids and Structures
270
DOI
出版状态已出版 - 15 5月 2023

指纹

探究 'Multiscale image-based modeling of progressive damage in 3D5D braided composites with yarn-reduction' 的科研主题。它们共同构成独一无二的指纹。

引用此