Overview: Recent advances and success stories in Additive Manufacturing (AM) or 3D printing have resulted in references to the process as the “next industrial revolution.” Producing parts without defects or that will adhere to failure standards are part of the challenge for the continued rapid growth of the industry. The variation in parameters during a build can contribute to the formation of defects. Most research approaches the problem from a material science perspective, but the large data quantities generated during AM builds naturally lend themselves to the use of analytics. We perform an initial investigation using a multi-pronged approach comprised of analytics and statistics for data discovery to identify areas for process improvement and promote the potential for advanced defect detection.
Significance and Impact: The large data quantities generated during AM builds naturally lend themselves to the use of analytics. The work indicates that multivariate statistics can aid in the identification of areas of concern within a part, leading to potentially more robust build processes and cost saving measures.
Sponsor/Facility: DOE/MDF, work was performed at ORNL