In-situ Melt Pool Monitoring of Laser Aided Additive Manufacturing using Infrared Thermal Imaging

In-situ monitoring is critical for detecting process anomalies and defect occurrences in additive manufacturing (AM). Traditional vision-based sensing approaches focused on extracting melt pool geometric information, while thermal-based monitoring focused on melt pool temperature measurement. This paper proposes an in-situ melt pool monitoring method utilising infrared thermal imaging for a robot-based direct energy deposition (DED) process. A high-resolution infrared thermal camera is employed to monitor the melt pool region, and a ROS-based multi-nodal software was developed to enable in-situ thermal image processing and melt pool feature extraction. The key contribution of this work is the development of a multi-feature extraction pipeline. Both melt pool geometric and thermal characteristics, such as contour area, centroids, elliptical width, peak temperature, and temperature variance, can be extracted and visualised in real time. The image processing and feature extraction pipeline can work concurrently with the sensor data acquisition. Experiment results are presented to show the effectiveness of the proposed in-situ melt pool monitoring method. It is found that melt pool geometric and thermal features share a similar trend in the temporal domain.