Framework for Curation and Distribution of Reference Datasets on the Example of Creep Data of Ni-Base Superalloys

This presentation introduces our current effort to develop a best practice framework for generating and distributing MSE reference datasets using creep data of Ni-based superalloys as an example, building on previous approaches [1,2]. Such reference datasets can be used in the verification of testing equipment or to compare experimental or simulation results among different laboratories or collaborative centers.

Experimental results can emerge, for instance, in the context of validating the creep behavior of newly developed alloys, by comparing their properties with the well-known behavior of a standard material. Simulation results can arise while performing simulation-based alloy design and optimization. We consider the data generation and usage process to be a cycle which can be divided into an upstream (data generation and distribution) and a downstream (data usage) component (see Figure 1). Our framework concept, for reference data, includes the generation of the data along with an agreement on the necessary metadata, considering the intended usage, and the integration into an easily accessible
digital infrastructure to annotate and discover/distribute the datasets using a metadata schema accordingly aligned to a domain ontology. ... mehrFurthermore, we aim to develop a widely accepted definition of reference data of materials, reach a self-supported community process, and motivate the discussion about reference material datasets and necessary metadata.

The digital infrastructure concept includes FAIR Digital Objects (FDOs) which are created to support machine readability and accessibility using Persistent Identifiers. With the help of an ontology, the reference dataset for creep data can be semantically described enabling the digital representation and reuse of shared concepts from domain experts, the community, and beyond. The interplay between a digital infrastructure, the agreement on necessary metadata and related metadata schema, and an underlying ontology will ensure that data and results can be easily shared, taken up, and re-used, thus fulfilling the FAIR principles (Findable, Accessible, Interoperable, Reusable).
The development of the intended framework is only possible through the collaboration of experts from materials science, semantics and ontologies, and data science, including data infrastructure service developers.

References
[1] Skrotzki, B., & Schriever, S., BAM reference data: results of ASTM E139 -11 creep tests on a reference
material of Nimonic 75 nickel-base alloy [Data set]. Zenodo, 2021, DOI: 10.5281/zenodo.5106606
[2] Eggeler, G., & Bürger, D., Creep Data for ERBO/1 (CMSX-4) in the range from 720-1080°C and 140-850 MPa
[Data set]. Zenodo, 2023, DOI: 10.5281/zenodo.7663974

Acknowledgments
Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under the National Research Data Infrastructure – NFDI 38/1 – project number 460247524.