Skip to content

Latest commit

 

History

History
113 lines (97 loc) · 5.16 KB

proposal.md

File metadata and controls

113 lines (97 loc) · 5.16 KB

Proposal

Motivation

Phase field models are a bridge between atomic-scale methods and the macro scale and, thus, a key component of ICME workflows. However, phase field simulations require extensive computational resources and generate large volumes of raw field data. A FAIRer, more systematic, approach to archiving phase field data has the potential to improve ICME workflows substantially. Currently, researchers are hindered in phase field research and engineering in all aspects of the FAIR paradigm with regards to both simulation and data archiving. In particular, developing better AI surrogate models for phase field could be improved with a more systematic approach to simulation management and data archiving.

In this project, we propose using a similar idea to the approach outlined by the Data Dictionaries Working Group. They state that "metadata that travels with the data and enrich dataset context by making relationships to other datasets in the materials research community explicitly". In particular, this approach enables researchers to query for datasets that are currently almost impossible to locate and reuse without working directly with publication authors. Currently, the PFHub project uses its own lightweight ontology to categorize and compare phase field results for the benchmark problems published by the CHiMaD phase field workshop participants. This takes the form of a small YAML file with links to relevant raw or processed tabular data. The data is archived with the YAML file alongside and the associated link is provided to the PFHub project. We propose a similar approach for archiving all phase field data outputs in general (not just those from the PFHub benchmarks).

The main aim of the working group is to provide an ontology or schema to describe phase field simulations and the corresponding data. This might include a simple link to a problem definition, a description of the numerical approach, computational resources used during the simulations, categorization of the physical model as well as links and descriptions of the data files. As an aside, it is worth noting that the metadata and provenance issues for storing simulation data effectively have not been widely addressed by the materials data community. Clearly, a successful approach to these issues within the phase field community could be disseminated into other sub-fields of materials science.

Work Plan

The main aim of the working group is to generate a lightweight ontology. However, this presents a small, but not insignificant hurdle for the researcher publishing phase field data. A secondary aim of the working group is to facilitate researchers in generating schema files via templates, web forms and command line tools. The main steps in the work plan are:

  1. Generate and finalize an initial schema

    • Use the existing PFHub LinkML repository as a basis for generating the schema.
    • Determine use cases (how would other researcher use phase field data).
    • Finalize a schema over first few months of WG

  2. Implement web tool, templates or command line to generate schema files

    • Develop a web tool for generating schema files similar to codemeta.json
    • Create prefilled examples of the YAML / JSON files.
    • Co-opt existing or use new command line tool to query repositories, ask questions, populate templates and push entries to Zenodo or similar service.
    • A possible long term goal would be a phase field registry similar to PFHub.

  3. Develop working examples

    • An example using an existing PFHub benchmark result and integrated into the PFHub registry.
    • An example using a study of numerical convergence using a series of simulations at varying discretizations.
    • An example phase field schema and associated data published in conjunction with a phase field publication.

  4. Approach wider phase field community for feedback

    • Present examples and tools at CHiMaD phase field workshops
    • Present same at the MaRDA meeting
    • Present at larger conference.
    • Publication describing schema and case study examples.

Goals and Expected Impact

  • Initial goal/impact would be to have the phase field schema used in conjunction with some phase field publications.
  • The PFHub project adopting the new schema.
  • Schema is widely used within the phase field community to publish results to generate FAIR content.
  • Secondary uses of the published phase field data to generate phase field AI surrogate models.
  • Other materials science communities adopting the standard for simulation data.

Deliverables

Within 18 months of the commencement of the working group, we will deliver:

  • A Working Group Note, published on the MaRDA Alliance website including some feedback from wider phase field community.
  • Implemented working examples as outlined in part 3 of the Work Plan above.