Multimodal Data Fusion and Analytics for the Additive Manufacturing Value Stream

Across current-day materials, processes and structures engineering (MPSE), research and development are bringing together all aspects of part design, materials, and manufacturing within common digital/simulation/data frameworks. Nearly all aspects of additively manufactured (AM) parts, especially for fused powder processes, innately operate from a digital basis. For these, the new challenge is to develop the infrastructures to fully utilize the data streams to inform design, select process/build conditions, enhance part inspection, and to support quality and certification systems. While AM processing provides designers with far more degrees of freedom and flexibility in geometric or topological design, the processes also introduce new levels of processing, material and microstructural complexity. Further, AM is a net-shape technology, and AM-produced parts might contain both intricate internal passages, and at the same time have complex outer geometries, that together create difficulties for current non-destructive inspection and quality assurance practices. New inspection and data analytics methods need to be advanced concomitantly with part design and manufacture. Ideally, one wants these protocols to reside within a systematic, generalized framework that makes use of the wealth of digital data associated with AM. Further, the new methods must deliver comparable or better probabilities of flaw detection as existing tools and be transferable across part geometries and materials types. Collectively, these requirements suggest directions for software and data management frameworks. DREAM.3D is an open-source software application for ingest, correction, analysis, reconstruction, instantiation, and archiving high-dimensional materials data operations over a hierarchy of scales. The current report shows progress and results from developing DREAM.3D for MPSE. Examples focus on fusing data from design, process monitoring, and post build inspection; performing analytics; and feeding back knowledge for process improvement and part quality engineering for AM.