Abstract This paper deals with the evaluation of several homogenization methods for short fiber-reinforced composites (SFRC) in isothermal linear elasticity. In order to have an unbiased assessment, only full-field finite element (FE) versions of the methods are developed and studied, no analytical mean-field models are considered. Firstly, the FE homogenization of unidirectional (UD) SFRC is performed using two computational models: representative volume elements (RVE) and unit cells, and comparing their predictions. Secondly, for the homogenization of misaligned SFRC, a model RVE seen as an aggregate of UD pseudo-grains (PG) is homogenized in two steps. In the first step, the effective response of PGs is obtained by FE analysis on UD unit cells. The second step is performed with one of three homogenization schemes: Voigt, Reuss and a FE-based version of the Mori-Tanaka (MT) model. The latter is equivalent to a direct MT homogenization of the RVE. The actual SFRC microstructure is 3D, but because of limited computer resources, the current article is restricted to 2D analyses, which nevertheless provide some guidance. A parametric study is conducted using several fiber volume fractions and orientation tensor components. Both the effective properties of RVEs and the mean strains inside individual fibers are assessed against reference results obtained by direct FE homogenization of the actual RVEs.

中文翻译：

---

短纤维增强复合材料：弹性中各种均匀化方法的无偏全场评估

摘要 本文涉及对短纤维增强复合材料 (SFRC) 等温线弹性的几种均质化方法的评估。为了进行无偏评估，仅开发和研究了方法的全场有限元 (FE) 版本，不考虑分析平均场模型。首先，使用两个计算模型执行单向 (UD) SFRC 的 FE 均质化：代表性体积元素 (RVE) 和单位单元，并比较它们的预测。其次，对于未对齐的 SFRC 的均质化，被视为 UD 伪晶粒 (PG) 聚集体的模型 RVE 分两步均质化。第一步，通过对 UD 晶胞的有限元分析获得 PGs 的有效响应。第二步使用以下三种均化方案之一进行：Voigt、Reuss 和基于有限元的 Mori-Tanaka (MT) 模型版本。后者相当于 RVE 的直接 MT 均质化。实际的 SFRC 微观结构是 3D，但由于计算机资源有限，当前文章仅限于 2D 分析，但仍提供了一些指导。使用几种纤维体积分数和取向张量分量进行参数研究。RVE 的有效特性和单个纤维内的平均应变都根据通过实际 RVE 的直接 FE 均质化获得的参考结果进行评估。使用几种纤维体积分数和取向张量分量进行参数研究。RVE 的有效特性和单个纤维内的平均应变都根据通过实际 RVE 的直接 FE 均质化获得的参考结果进行评估。使用几种纤维体积分数和取向张量分量进行参数研究。RVE 的有效特性和单个纤维内的平均应变都根据通过实际 RVE 的直接 FE 均质化获得的参考结果进行评估。