Strength Prediction Sensitivity of Foamed Recycled Polymer Composite Structures due to the Localized Variability of the Cell Density Distribution

• Main Authors: Daniel P. Pulipati, David A. Jack
• Format: Article
• Language: English
• Published: MDPI AG 2020-07-01
• Series: Journal of Composites Science
• Subjects: composite foams; closed cells; image processing; finite element analysis
• Online Access: https://www.mdpi.com/2504-477X/4/3/93

The need for novel methods for the reuse of post-industrial/post-consumer polymer solid wastes (PSW) is of increasing societal importance. Unfortunately, this objective is often limited due to material stream variability or insufficient load-carrying capacity of the fabricated goods. This study investigates a large format fiber-reinforced structural member that contains spatially varying material properties, specifically density. The application is focused on the unique features of closed-cell foamed composite structures made from recycled post-industrial/post-consumer PSW composed of High-Density Polyethylene (HDPE) and Glass Fiber Polypropylene (GFPP). The structures in this research are manufactured using a hybrid extrusion process, which involves foaming enabled by chemical blowing agents that form a fully consolidated solid outer shell and a closed-cell core. The cell distribution is inhomogeneous, in size distribution and spatial distribution, leading to significant spatial variations of the local effective stiffness. To understand the correlation between density variations and effective stiffness and strength, a low-cost method using digital imaging is introduced and integrated into a finite element subroutine. The imaging approach includes sectioning the structural member and analyzing the resulting image using various custom imaging processing techniques in the MATLAB environment. The accuracy of the imaging technique was experimentally verified using a Keyence digital microscope, and the error was found to be 3% in any given spatial feature. The processed image is then correlated to a localized density map of the cross-section using a weighted spatial averaging technique, and the local effective material properties of the foamed region are predicted using the presented micromechanical approach. The local stiffness is a function of void density, local fiber orientation, constitutive behavior of both the fiber and the matrix blend, and the non-linear response of the matrix blend. The spatially varying stiffness and nonlinear strength response at each spatial location are then integrated into a finite element subroutine within the COMSOL multiphysics environment, and results are presented for the deflection and internal stress state of the composite structure. Results indicate that the internal microstructural variations have a nominal impact on the bulk deflection profile. Conversely, results show the peak of the internal stress is increased by ∼11% as compared to the uniform core assumption, thus safe designs must consider core density spatial variations in the final product design.