Toward a Standard Data Architecture for Additive Manufacturing

Abstract

To advance additive manufacturing (AM), a scalable architecture is needed to structure, curate and access the data from AM R&D projects that are conducted to evaluate new materials, processes and technologies. Effective project metadata management enables the sharing of AM domain knowledge. This work introduces an AM data modeling architecture to capture pedigree information from AM projects which enables the traceability of the material. This overall AM model includes five modules covering information about (1) project management, (2) feedstock materials, (3) AM building and post processing, (4) microstructure and properties measurements and (5) computer simulations. The objective of this design is to ease the integration of the heterogeneous datasets from different sources and allow for extensions, for example, to incorporate sub-models from other efforts. As a proof of concept, the material and process models defined in the paper capture the major metadata elements for laser powder bed fusion AM. To demonstrate the effectiveness of the architecture, the models are implemented using extensible markup language and preliminarily tested using the project data from America Makes. Additional data sub-models can be integrated in this architecture without affecting the existing structure.