Overview

Overview#

The OpenMC Fusion Benchmarks (OFB) project uses a structured, YAML-based specification format to define every component of a radiation transport benchmark. This approach ensures clarity, reproducibility, and automation in defining, running, and analyzing benchmarks.

Each benchmark is defined by a specifications.yaml file that is validated against a strict schema (benchmark_schema.yaml) to enforce consistency and completeness across the benchmark suite. We define a benchmark specifications as:

The minimum amount of technical data necessary to unambiguously model a benchmark and collect standardized results.

Specifications Sections#

The specifications.yaml file captures all essential aspects of a benchmark, including:

  • Metadata
    General information such as benchmark name, description, references, authors, and version.

  • Materials
    Composition, temperature, density, and other nuclear properties defined in a structured format.

  • Geometry
    CAD-based geometry definitions, including references to CAD files and meshing parameters.

  • Sources
    Neutron or photon source definitions, including energy, spatial, and angular distributions.

  • Settings
    Transport code configuration such as particle count, batches, and other run control parameters.

  • Tallies
    Definition of observables (e.g., flux, dose) with filters and expected data structure.

  • Uncertainty Quantification (Optional)
    Setup for input perturbations, sampling, and uncertainty propagation metrics.

  • Irradiation Schedule (Optional)
    Time-dependent irradiation and cooling steps for activation and shutdown dose rate analysis.

Schema Validation#

To guarantee interoperability and catch user errors early, every specifications.yaml file must conform to the benchmark_schema.yaml. The schema:

  • Enforces required fields and correct types

  • Validates units, formats, and structure

  • Allows custom extensions while preserving core validation

Validation is automatically handled by the OFB Python API.

Why This Matters#

  • Consistency across all benchmarks

  • Automation of modeling, simulation, and analysis workflows

  • Comparability of results across codes and experiments

  • Modularity to support method testing and rapid development

Learn More#