Investigation of additively manufactured PEEK spur gears reinforced with graphene and natural fibers using hybrid AI techniques

High-temperature hydrogen production reactors have to use spur gears with high mechanical, thermal, and a good level of durability. This paper offers an original combined design and optimization model of hybrid additively manufactured PEEK-based composites, manufactured by incorporating graphene platelets and flax fibers. The samples and spur gears are fused through Fused Filament Fabrication to form filaments made of graphene (1–5 wt. %) and flax fibers (5–20 wt. %), respectively. Different mechanical characteristics, i.e., tensile strength, elastic modulus, elongation at break, and resistance to heat, are studied. To determine microstructural characteristics and dispersion of reinforcements, Scanning Electron Microscopy (SEM) was used. Response Surface Methodology (RSM) was used to model and optimally control the interacting material composition and processing parameters. The data obtained was then used to train hybrid Artificial Intelligence (AI) models to be able to predict their performance accurately. It was found that the addition of reinforcements enhances strength and thermal stability and decreases ductility. The composite with 3 wt% graphene and 12.5 wt% flax fiber showed the highest overall performance with a high level of material-level potential of spur gears and similar mechanical components in high-temperature hydrogen production settings, although additional hydrogen exposure and tribological validation is needed before this material can be used in-service.