Abstract

Advancements in additive manufacturing have enabled the creation of intricate ceramic structures with complex geometries. Among these, extrusion free-forming has emerged as a versatile and cost-effective method for fabricating ceramics used in both artistic and engineering applications. This study explores the integration of ChatGPT in Socratic-style dialogue to guide and troubleshoot the 3D printing process of barium hexaferrite via extrusion free-forming. Ceramic pastes were prepared by dispersing mixed precursor powders in aqueous solutions containing poly(vinyl alcohol) (PVA) as a binder, poly(ethylene glycol) (PEG-400) as a plasticizer, and EFKA^®FA4620 as a dispersant. The paste was then extruded using a customized 3D printer equipped with a syringe-based delivery system. ChatGPT was employed to support various stages of the workflow, including prompt-based troubleshooting of nozzle clogging, addressing cracking in printed bodies, and optimizing paste formulation. One notable intervention guided by ChatGPT involved reducing the PEG-400 content from 317 mg to 300 mg, which significantly improved paste consistency, printability, and the overall quality of the printed barium hexaferrite structures. This study highlights the practical value of ChatGPT in assisting real-time problem-solving in ceramic additive manufacturing. The findings underscore the potential for integrating large language models into experimental design and materials processing, promoting collaborative learning and iterative optimization among engineers, scientists, and artists.