Nonparametric Bayesian framework for material and process optimization with nanocomposite fused filament fabrication

Abstract

The convergence of polymer nanocomposite development and additive manufacturing creates unprecedented opportunities for advancing novel materials design and product customization in various industrial applications. Graphene has been used as fillers to enhance the host polymer materials for additive manufacturing. However, creating and additively manufacturing a new polymer nanocomposite is challenging due to the strong interactions between graphene concentration and process parameters. This work aims to develop a data-driven nonparametric Bayesian framework to unify nanocomposite design and part fabrication by integrating statistical learning and optimization into quality prediction and process optimization. It regulates the design and fabrication of different ABS/graphene nanocomposite specimens using a fused filament fabrication (FFF) process, and ensures successful printings and accurate quality prediction while maintaining its design flexibility and customization. The preliminary experimental results showed that adding graphene to ABS (5 wt%) and adjusting FFF process parameters significantly improved the surface roughness of the printed FFF specimens and increased their ultimate tensile strength by 25–25%. With surface roughness as the quality indicator, Gaussian process regression in the proposed framework achieved surface roughness modeling with high prediction accuracy (MAPE = 0.13), and Bayesian optimization quickly discovered the optimal process parameters within merely 5 iterations. Compared to trial-and-error or computational approaches, the data-driven framework can bypass the complexity of physics-based models, explore the feasible space efficiently, and reduce time and cost for ABS/graphene nanocomposite design and FFF fabrication. It can be easily extended to other nanocomposites and will have a profound influence on advancing their applications and innovations in various industries.