In-situ monitoring of Material Extrusion processes via thermal videoimaging with application to Big Area Additive Manufacturing (BAAM)

• Published in: Additive manufacturing Vol. 58; no. C; p. 102995
• Main Authors: Caltanissetta, Fabio, Dreifus, Gregory, Hart, Anastasios John, Colosimo, Bianca Maria
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.10.2022; Elsevier
• Subjects: Big Area Additive Manufacturing; Composites; Extrusion additive manufacturing; In-situ monitoring; Thermography; Composites; In-situ monitoring; Big Area Additive Manufacturing; Extrusion additive manufacturing; Thermography
• ISSN: 2214-8604; 2214-7810
• DOI: 10.1016/j.addma.2022.102995

Material Extrusion (ME) is becoming one of the most promising class of Additive Manufacturing (AM) processes in the manufacturing panorama for its capability of realizing products with complex shapes, potentially in large dimensions, for a variety of applications and materials for different industrial sectors. Despite the significant advances in material range and machine capabilities, the lack of process stability and repeatability of ME still represents an obstacle for the adoption of this process on large industrial scales. Following the success of in-situ monitoring in other AM technologies, some studies are investigating the possibility of embedding in-situ sensors on commercial machines. Among them, thermal infrared monitoring systems have gained lots of interest for the crucial role of temperature evolution in final part quality and mechanical performance. This study presents a data-driven methodology for in-situ monitoring of thermal profiles. The proposed algorithm extracts, models, and compares the cooling history at different locations. By combining a tracking algorithm and automatic detection of region of interests and robust regression for functional data modeling, we propose a solution to study the cooling history of the 3D printed geometry as the process evolves, to draw general conclusions on the process and detect anomalous thermal profiles, which may entail lack of bonding or other flaws. This paper further proposes a novel approach to automatically detect local temperature inhomogeneities, also known as “hot” and “cold spots”, which are often symptomatic of critical defects in the 3D printed part. In-situ detection of hold- and cold-spots can be particularly useful for large scale products or functional products printing. The proposed methods can be applied to any ME process, but it is specifically applied to a real case study of Big Area Additive Manufacturing (BAAM), in collaboration with the Oak Ridge National Laboratory. Results show the effectiveness of all the proposed solutions and pave the way towards zero-defects production.