Describing the Material Behavior of Steel and Carbon Fiber Reinforced Composites Using a Combined Damage-Plasticity Approach

Abstract

Metal fiber hybrids (MFH) exhibit outstanding mechanical properties. They combine the advantages of ductile metallic materials with the well-known advantages of classical glass or carbon fibers in polymer matrices. Previous research has shown that these hybrid material concepts can improve structural integrity and energy absorption while maintaining their excellent weight-specific mechanical properties as well as allowing a wider range of multifunctional applications. In today’s component design process, simulation is a powerful tool for engineers to exploit the full mechanical potential of the material used. However, describing the material behavior including its multifunctional usability in numerically aided design processes of components is currently one of the major challenges for MFH. Against this background, this work focuses on the development and evaluation of a description method for MFH in the finite element analysis (FEA). A steel and carbon fiber reinforced epoxy resin (SCFRP) with hybridization at the laminate level is chosen as the reference material. To describe the behavior of unidirectional steel fiber reinforced plastic (SFRP) layers, a material model combining an orthotropic damage model and a 1D-plasticity model is proposed and implemented as a user-defined subroutine for LS-Dyna. In addition, SCFRP laminates are manufactured, tested under tensile loading, and used to parameterize the material models and to validate the description method for SCFRP. In this study, it is shown that the description method in combination with the newly developed material model is able to describe the complex failure mechanism of SCFRP. In particular, with respect to the material behavior up to the failure of the carbon fibers, a very good mapping accuracy can be achieved. Strain localization effects occur in both numerically predicted and experimentally measured post-failure behavior. Therefore, it could be concluded that the accuracy of the numerical predictions strongly depends on the geometric resolution of the discretization.