A Mean-Field Homogenization Method-Based Design of Experiments for Fiber-Reinforced Plastics | NWC23

Abstract

Fiber-reinforced plastic (FRP) materials are widely used in our daily life, from consumer packages to home appliances, from automobiles to commercial airplanes, just to name a few. FRPs are used as a lighter alternative to metals, with lower costs, comparable durability, better corrosion resistance, and better manufacturing versatility.

To increase the strength and durability of plastic parts, chopped fibers are added to the resin as stiffeners when parts are injection molded. When the injection gates and runner system are not carefully designed, manufacturing defects can be introduced. If parts are not properly cooled down before being ejected out of the mold, significant residual stress can rise and cause parts to warp. In addition, the mechanical properties of FRPs often exhibit a strong anisotropy and temperature-dependency, which largely depends on how fibers are oriented during thermal setting. Understanding the effects of geometric and manufacturing parameters on the material responses of the molded part under service loads is crucial to finding the best design to meet performance requirements. These issues pose unique challenges to the design and manufacturing of FRPs.

Previously, a fully integrated workflow was developed for modeling FRPs with mean-field homogenization (MFH) method-based multiscale models. Automated through a simulation process, this workflow starts with part design and includes an injection molding simulation followed by a structural finite-element analysis. The fiber orientation distribution predicted by the injection molding simulation is used in the MFH-based multiscale models to accurately characterize the mechanical behaviors of FRPs.

In this study, we extended this workflow to design of experiments to explore design alternatives. Using a plastic mounting boss as an example, the effects of the part thickness and injection gate location on the maximum von Mises stress in service have been studied. An optimal design has been obtained by minimizing the stress and weight together.


Design parameters (d=gate location offset, t=thickness) and plastic injection simulation result (arrows indicate fiber orientation).


Structure simulation definitions


Simulation process for DOE analysis. The two gear icons represent the plastic injection simulation and structure simulation, respectively.