Similarity analysis for thermal signature comparison in metal additive manufacturing

• Published in: Materials & design 2022-12, Vol.224 (C), p.111261, Article 111261
• Main Authors: Chandrasekar, Sujana, Coble, Jamie B., List, Fred, Carver, Keith, Beauchamp, Serena, Godfrey, Amy, Paquit, Vincent, Babu, Sudarsanam S.
• Format: Article
• Language: English
• Subjects: Additive Manufacturing; In-situ monitoring; Infrared thermography; Similarity analysis; Thermal signature
• ISSN: 0264-1275; 1873-4197
• DOI: 10.1016/j.matdes.2022.111261

[Display omitted] •An in-situ monitoring method and similarity analysis algorithm has been developed as a potential tool for part qualification based on thermal signature comparison in additive manufacturing.•The similarity analysis algorithm combines shape and magnitude similarity in a manner unique to additive manufacturing.•Using similarity analysis, variations in thermal signatures associated with change in part geometry have been demonstrated.•Comparison with microstructure characterized using ex-situ microscopy indicated that similarity analysis correlates well with grain size and microstructure variation within a part.•A proof of concept is also presented to demonstrate similarity analysis to identify anomalous regions corresponding to potential defects in the part. Metal additive manufacturing differs from traditional manufacturing processes primarily due to repeated heating and cooling cycles that are inherent to the process. Such cycling leads to different thermal signatures and associated heterogeneities in solidification, solid-state transformation, microstructure, and properties of the printed part. There is, therefore, a need to develop methods to compare thermal signatures in a part with reference to part qualification criteria. In this work, we have developed and demonstrated a similarity analysis procedure to meet this need. Thermal signatures are extracted from in-situ infrared (IR) thermography during a part built using a laser powder bed fusion (l-PBF) machine. Layer-wise variations in thermal signature as a function of part geometry were revealed. Ex-situ microstructure characterization was used to validate the results from similarity analysis. Results and analysis confirmed that our similarity analysis is indeed a useful tool for rapid comparison of thermal signatures, i.e., heterogeneity within the part. Similarity analysis was also useful in identification of anomalous thermal signatures. The analysis method paves the way for deploying in-situ monitoring of AM parts for qualification of AM components.