1 Proceedings of the ASME 2017 International Design Engineering Technical Conferences & Computers & Information in Engineering Conference IDETC/CIE 2017 August 6-9, 2017, Cleveland, Ohio, USA IDETC201768457 A Collaborative Data Management System for Additive Manufacturing Yan Lu, Paul Witherell, and Alkan Donmez National Institute of Standards and Technology, Gaithersburg, Maryland 20899 Email: [yan.lu, paul.witherell, alkan.donmez]@nist.gov ABSTRACT As additive manufacturing (AM) continues to mature as a production technology, the limiting factors that have hindered its adoption in the past still exist, for example, process repeatability and material availability issues. Overcoming many of these production hurdles requires a further understanding of geometry- process-structure-property relationships for additively manufactured parts. In smaller sample sizes, empirical approaches that seek to harness data have proven to be effective in identifying material process-structure-property relationships. This paper presents a collaborative AM data management system developed at the National Institute of Standards and Technology (NIST). This data management system is built with NoSQL (Not Only Structured Query Language) database technology and provides a Representational State Transfer (REST) interface for application integration. In addition, a web interface is provided for data curating, exploring, and downloading. An AM data schema is provided by NIST for an alpha release, as well as a set of data generated from an interlaboratory study of additively manufactured nickel alloy (IN625) parts. For data exploration, the data management system provides a mechanism for customized web graphic user interfaces configurable through a visualization ontology. As a collaboration platform, the data management system is set to evolve through sharing of both the AM schema and AM development data among the stakeholders in the AM community. As data sets continue to accumulate, it becomes possible to establish new correlations between processes, materials, and parts. The functionality of the data management system is demonstrated through the curation and querying of the curated AM datasets. 1 INTRODUCTION Additive manufacturing (AM) is a process capable of building up complex shapes layer upon layer directly from a three- dimensional digital representation. As a production technology, AM minimizes the time lag between design and realization of a part while enabling novel design opportunities. However, since both the shape and the material properties of parts are formed during the AM process, manufacturers and end users have difficulty in predicting whether parts will meet specifications (shape, surface and material properties) and ensuring repeatability, consistency, and reliability across different AM machines and locations. Qualification of additively manufactured parts is one of the most serious hurdles to the broad adoption of additive manufacturing [1]. A key step in overcoming these hurdles is to characterize AM materials and geometry- -process-structure-property relationships. To develop such relationships, sufficient information about feedstock materials, processes, machines, and final part performance tests must be collected and analyzed accordingly. Such an effort requires a well-populated database containing the necessary pedigreed information. This task is difficult and extremely expensive for any single company, especially given the limitation of currently available standard methods and protocols for AM feedstock material characterization and AM part property testing [2]. Several data system development efforts have been reported for AM. For example, the Senvol Database [3] provides researchers and manufacturers with open access to industrial AM machines and materials information. A search on the Senvol Database finds properties of most industrial additive manufacturing machines and materials, and information on their compatibility. Granta, a material information management technology provider, has collaborated in several European Framework Seven projects in the field of material information management. The results from those projects are not currently available for public access, and the underlying data structure ('schema') is proprietary to Granta users. The Materials Selection an Analysis Tool (MSAT) database led by Department of Defense (DOD) was built based on GRANTA:MI [4], which captured some data from NASA and it is available to approved users[5]. None of the existing AM databases are publicly accessible, and none provide a sufficient amount of reliable material property data for much of the necessary AM analytics. These two hurdles are significant barriers to adoption of this emerging transformative technology as a viable production alternative to traditional manufacturing methods, especially for small manufacturers. To obtain the multitude of data sets rich enough for AM geometry-material-process–structure-property relationship identification, the full processing history of thousands of samples