Abstract

Fiber-reinforced composites are designed to withstand extreme conditions possibly for decades. To extract the fibers from the polymer matrix, the recycling is typically
conducted under conditions that deteriorate the recycled fibers’ properties, such as tensile properties and surface quality of the fibers. A further issue with recycled fibers is their difficult manageability, as the fibers do not stay together, which can make the production of industrial-scale second-generation fiber reinforcements impossible.

This work aims to find a polymer-based binder system that allows thermo-chemically recycled glass fibers to be further processed into a nonwoven fabric that has commercial potential. The goal is for the produced nonwoven fabric to have interfacial properties compatible with an epoxy matrix, ensuring efficient load transfer from the matrix to the fibers. Additionally, the nonwoven fabric should be easily manageable to keep its fibers together.

Recycled fibers were treated either with a sizing and a binder or a binder alone to assess the necessity and effect of the sizing treatment. The molecular weight, the amount of binder on the fiber surface, and different binder chemicals were evaluated. Micro-scale interfacial shear strength was measured using a microbond apparatus. The macro-scale interphasial properties formed by the nonwoven fabric with the epoxy matrix were examined using dynamic mechanical analysis. Flexural resistance measurements for the nonwoven fabric enabled the evaluation of its processability. Thermogravimetric analysis and stereo microscope images were taken of the nonwoven fabrics to determine the amount of binder and the binding method.

The results indicated that the significance of molecular weight and the effect of the  amount of the binder decreases when sizing is used. The epoxy binder plasticized the interphase at a sufficiently high concentration, which weakened the load transfer, but sizing reduced this effect. The polyurethane binder decreased the interfacial shear strength compared to sized fibers but by increasing the amount of the binder from 10 weight percent to 30 weight percent did not affect the interfacial properties in micro or macro scale tests. At both scales the polyurethane binder behaved most systematically with the epoxy matrix. The polyurethane binder proved to be the most suitable for the production of nonwoven due to its manageability and sufficient surface properties. The binder formed a web structure on the fibers, which allowed the fibers to stay together and be easily handled