Session: 16-01-04: Mechanical Performance III

Paper Number: 173405

Accessing Additive Manufacturing Benchmark Data for Analysis and Computation 

A primary objective of the Additive Manufacturing Benchmark Test Series (AM-Bench) effort is to make experimental additive manufacturing benchmarking data available in ways that facilitate use and reuse beyond the initial challenge problems.  While the data systems are comprised of a number of components that work together in complementary ways, described at https://nist.gov/ambench, this presentation will focus on how AM-Bench data can be accessed using two methods: 1) within the SciServer collaborative computational environment, and 2) via the NIST Science Data Portal application programming interface (API).  An overview with examples will also be given of how to access the data with appropriate metadata for both of these cases, as well as how the data relate back to the original NIST sources.

SciServer (https://sciserver.org) is a platform for server-side data analysis that can accommodate very large datasets, and copies of publicly released AM-Bench data are hosted there to facilitate access.  Having data close to compute allows AM-Bench data to be accessed in a way that can readily be used for analysis and modeling within the collaborative computing environment.  We will describe some of the features of the SciServer platform, what AM-Bench data are available, how the data are organized, which metadata are available about the datasets, and how the data can be accessed within the shared data volumes.

The primary location of the standalone AM-Bench datasets is the NIST Science Data Portal (https://data.nist.gov), which includes support for search, subject-related keywords, and topic filtering capabilities to facilitate the discovery of standalone datasets published by NIST.  The portal resources can be accessed both via a human-friendly user interface and an application programming interface.  We will give examples of some AM-Bench records available in the Science Data Portal, along with information about the related metadata for the records and how to access it.  The Scientific Data Portal API allows access to the data from a range of computing resources, making scripted use and analysis easier, which supports the objective of facilitating AM-Bench data use and reuse.

Finally, we will describe how these individual components of the data management system can be linked via scripts with the additional metadata located in the AM-Bench metadata catalog in order to aggregate and analyze a more complete set of benchmarking data.  We will consider some details of implementation and also discuss how these components of the data management systems fit into the larger AM-Bench plans and future efforts.

Presenting Author: Chandler Becker Research Data and Computing Office, National Institute of Standards and Technology

Presenting Author Biography: Dr. Chandler Becker is a Materials Research Engineer at the National Institute of Standards and Technology. She has recently worked mostly on projects related to materials data availability, and she is currently a member of the data management team for the Additive Manufacturing Benchmark Test Series (https://www.nist.gov/ambench). Dr. Becker also works on application of data science approaches with an emphasis on techniques for analytic reproducibility, documentation, sharing, and reusability.

Authors:

Chandler Becker Research Data and Computing Office, National Institute of Standards and Technology
Gretchen Greene Research Data and Computing Office, National Institute of Standards and Technology
Gerard Lemson Institute for Data-Intensive Engineering and Science, The Johns Hopkins University
Lyle Levine Materials Science and Engineering Division, Material Measurement Laboratory, National Institute of Standards and Technology
Benjamin Long Software and Systems Division, Information Technology Laboratory, National Institute of Standards and Technology
Jaiwon Kim Institute for Data-Intensive Engineering and Science, The Johns Hopkins University
Shengyen Li Systems Integration Division, Engineering Laboratory, National Institute of Standards and Technology