|About this Abstract
||2021 TMS Annual Meeting & Exhibition
||Additive Manufacturing for Energy Applications III
||Toward Part Qualification: Thermal Signature Analysis Using Wavelet Transform in Metal Additive Manufacturing
||Sujana Chandrasekar, Jamie Baalis Coble, Amy Godfrey, Serena Beauchamp, Fred List, Vincent Paquit, Sudarsanam Suresh Babu
|On-Site Speaker (Planned)
A key challenge in increased adoption of Additive Manufacturing for critical energy applications is part qualification for different load conditions. Additively manufactured parts exhibit variation in material properties due to inherent variation in thermal process cycles, associated primarily with scan strategy and part geometry. To qualify parts, it is essential to identify regions having similar thermal cycles and those that are anomalous. We develop an in-situ monitoring method for the laser powder bed fusion process using infrared (IR) data collected during the build process. Our monitoring algorithm is built on the wavelet transform and used to identify similar thermal cycles based on IR data. Wavelet transform enables multiresolution analysis and can be used both for pattern recognition and anomaly detection. In-situ monitoring give us the ability to identify thermally similar regions based on scan strategy and also identify overheated regions. This approach is a promising step toward data-driven part qualification.
||Additive Manufacturing, ICME, Machine Learning