Spatio-temporal Analysis of Thermal Profiles in Extrusion-based Additive Manufacturing

Extrusion-based Additive Manufacturing (AM) processes have recently gained increasing attention in the scientific and industrial communities because of the wide range of processible materials (from thermoplastics to composite and biomaterials), printable volumes, and industrial applications. As for many other AM processes, the actual problems with process stability and repeatability are still limiting the industrial process adoption, as these problems can significantly impact on the final part quality. In this framework, a latest research trend aims at developing in-situ monitoring solutions for inline defect detection, in a zero-waste production perspective. Among the existing in-situ sensing techniques, many studies showed that in-situ thermography represents a viable solution to describe the temperature dynamic and validate the thermal models but very few approaches have been proposed to quantitively study the temperature evolution to quickly detect process instabilities. This paper presents a new approach to quickly analyse the temporal dynamic of temperature in the printed layer while providing a spatial mapping of the temperature homogeneities. Compared with previous methods, the current one has the main novelty feature of combining both the spatial and temporal signature in a synthetic mapping that allows to detect unstable or unusual problems. In order to show the effectiveness of the proposed solution, a real case study of Big Area Additive Manufacturing (BAAM) for composite materials is considered. The study shows that the provided method can clearly enhance defect detection and represents a new solution for detecting anomalous areas where thermal profiles behave differently with respect to the surrounding areas. The same methodology underlined the thermal evolution complexity in the BAAM case study and enabled the detection of local flaws, i.e., hot and cold spots.