Session: 12-04-01: Multiscale Models and Experimental Techniques for Composite Materials and Structures

Paper Number: 96784

96784 - Towards Verification and Validation of Modeling Dyneema Using the Embedded Finite Element Method 

Fiber reinforced composites, such as Dyneema, have been of interest to the ballistics community for some time due to their light weight and high tensile strength. The composites usually consist of 70-85% high strength fibers by volume held together by a polyethene matrix. There is an ongoing effort to create an accurate and efficient finite element model to predict the material deformation for armor design. A majority of this work has been focused on using microscale models unit cells of fiber and matrix material to inform the properties of macroscale models where the fiber and matrix materials are combined into an orthotropic material to represent the overall behavior. A different approach is being explored to maintain the independent material properties of the fibers and the matrix. Here an embedded element model of Dyneema, where truss elements representing bundles of fibers are embedded in a matrix material, is compared to experimental data from the literature. By incorporating the high strength fibers into a finite element model in a more explicit way, we hope to create a method of modeling that can easily be used on multiple length scales, and capture unique effects created by the layering of the fibers.

The embedded element approach is a method of finite element analysis where one element mesh is placed inside of a separate host  mesh. The elements’ deformations are coupled together so the embedded elements act as an inclusion in the host. This can be used to model composite materials, such as fiber reinforced composites. The embedded element method provides a unique type of macroscale model where the matrix and fiber components can be represented by two different element types, fibers as truss elements and the matrix as continuum elements, making the two components easily distinguishable while keeping the meshing process simple. When using the embedded element method, there is an inherent volume redundancy, as the two meshes occupy the same volume. This issue must be addressed in order to produce accurate results. In this work an explicit finite element code that removes the redundancy is shown.

By using denser embedded element distributions to represent Dyneema fibers, various deformation mechanisms are more clearly represented when comparing to experimental data. Additionally, by eliminating volume redundancy, the material properties used will be closer to the true properties of the fiber and matrix materials while still producing accurate predictions of material behavior. This is further explored by comparing experimental and computational results.

Presenting Author: Valerie Martin The Pennsylvania State University

Presenting Author Biography: Valerie Martin is a graduate student studying mechanical engineering at Penn State University. Her work focuses on finite element methods, composite modeling, and high strain rate impacts.

Authors:

Valerie Martin The Pennsylvania State University
Thomas Hannah The Pennsylvania State University
Stephen Ellis Los Alamos National Laboratory
Reuben Kraft The Pennsylvania State University