DIRECT FIBER MODEL VALIDATION: ORIENTATION EVOLUTION IN SIMPLE SHEAR FLOW
DIRECT FIBER MODEL VALIDATION: ORIENTATION EVOLUTION IN SIMPLE SHEAR FLOW
复制标题
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
S. Simon;A. Senior;T. Osswald
中科院分区：
文献类型：
--
作者：
S. Simon;A. Senior;T. Osswald
直接颗粒模型是预测纤维增强复合材料微观结构性能的一种很有前途的工具。为了验证我们的纤维取向预测的建模方法，在滑板流变仪中对压缩模塑增强聚丙烯样品进行了简单剪切流动。微计算机层析成像被用来测量多达60个剪切应变单位的取向张量。在粒子模拟中，采用了零剪切应变下的完全表征的微观结构来再现初始条件。将纤维放置在周期性的边界单元中，并应用与实验相匹配的流场。用建议的压缩模塑技术制作的样品显示出可重复的和可控制的初始取向。该模型与稳态取向具有较好的一致性，但在剪切过程开始时，取向演化较快。前言模拟纤维增强复合材料加工过程的计算工具已成为汽车工业不可缺少的工具。准确预测成型部件的微观结构的能力不仅是设计计算的关键因素，也是在模具制造之前解决收缩和翘曲等问题的关键因素。几乎所有商业实现的模型(Folga-Tucker、各向异性旋转扩散(ARD)减缩应变闭合(RSC))都使用实验确定的拟合参数。然而，这些实验成本高、耗时长，而且能提供的信息量有限。过去，粒子水平模拟被用来帮助理解这些类型的系统[1，2]。在这些模拟中，每根纤维都被单独建模为一系列刚性元件。纤维被放置在预定的流场中，并计算流体动力以及纤维与纤维的相互作用，以预测纤维的运动[3，4]。用数值方法获得参数比用实验方法获得参数更有优势。在数字设置中，所有参数都可以精确控制，详细信息始终可用，并且执行模拟的成本相对较低。这些模型的另一个优点是通过对单个纤维的实际运动进行建模而达到的高精度。本文的目的是在定义良好的简单剪切流中提供可靠的纤维取向演化数据，以帮助验证和开发用于增强纤维的多颗粒模型。选择简单剪切是因为它是大多数聚合物加工过程中存在的基本流动条件之一。选择它可以使我们直接将变形率与填充物的行为相关联。按照Cieslinski等人的方法，在滑动板式流变仪中对模压玻璃纤维增强的聚丙烯样品进行剪切。[6]。正如同一作者所指出的那样，模压成型不是一种合适的样品制备方法，因为它不能控制纤维的平面取向。因此，在这项工作中，我们将提出一种压缩成型技术，确保剪切实验中可控和可重复的初始纤维取向。将仿真结果与实验结果进行了比较。
Direct particle models are a promising tool for predicting microstructural properties of fiber reinforced composites. In order to validate our modeling approach for fiber orientation prediction, compression molded reinforced Polypropylene samples were subjected to a simple shear flow in a Sliding Plate Rheometer. Micro computed tomography was used to measure the orientation tensor for deformations up to 60 shear strain units. The fully characterized microstructure at zero shear strain was used to reproduce the initial conditions in the particle simulation. Fibers were placed in a periodic boundary cell and a flow field matching the experiment was applied. Samples created with the proposed compression molding technique showed repeatable and controlled initial orientation. The model showed good agreement with the steady state orientation; however, it showed a faster orientation evolution at the start of the shearing process. Introduction Computational tools to simulate the processing of fiber reinforced composites have become indispensable for the automotive industry. The ability to accurately predict the microstructure of molded components is a key factor not only for design calculations but also for addressing issues such as shrinkage and warpage before mold fabrication. Almost all commercially implemented models (Folgar-Tucker, Anisotropy Rotary Diffusion (ARD) Reduced Strain Closure (RSC)) use experimentally determined fitting parameters. However, these experiments are costly, lengthy and limited in the amount of information they can provide. Particle level simulations have been used in the past to aid in the understanding of these type of systems [1, 2]. In these simulations, each fiber is modeled individually as a chain of rigid elements. Fibers are placed in a predetermined flow field and hydrodynamic forces, as well as fiber-fiber interactions, are computed to predict fiber motion [3, 4]. Obtaining parameters numerically has advantages over obtaining them experimentally. In a numerical setup, all parameters can be accurately controlled, detailed information is always available and the simulations are relatively inexpensive to perform. An additional advantage of these models is the high accuracy that is reached by modeling the actual motion of individual fibers [5]. The objective of this paper is to provide reliable fiber orientation evolution data in a welldefined simple shear flow to aid in the validation and development of a multi-particle model for reinforcing fibers. Simple shear was chosen since it is one of the fundamental flow conditions present in most polymer processes. Selecting it allows us to directly correlate the rate of deformation with the filler’s behavior. Compression molded glass fiber-reinforced polypropylene samples were sheared in a Sliding Plate Rheometer following Cieslinski et al. [6]. As has been shown by the same author, compression molding is not a suitable sample preparation method since it has no control over the planar orientation of the fibers. In this work, we will therefore present a compression molding technique which ensures a controlled and repeatable initial fiber orientation for shear experiments. Results from both simulation and experiment are compared.