control_packages.json

{
  "schema_version": "fibre_control_packages.v2",
  "visibility": "evaluator_reference_only",
  "rule": "A single episode may be used for an operations curriculum; a selected control package requires all of its conditions and paired records, without cherry-picking successful cases",
  "control_packages": [
    {
      "id": "C_CLADDING_INTEGRITY",
      "branch_ids": [
        "GE_SILICA_CRACK",
        "RELEASE_GE"
      ],
      "comparison_and_qualification": "Ge/silica cladding cracks and released fragments versus the released Ge/ASG core; process parameters are missing, so do not claim strict quantitative single-variable attribution"
    },
    {
      "id": "C_CAPILLARY",
      "branch_ids": [
        "GE_BSG_BREAKUP",
        "GLASS_GE_ASG"
      ],
      "comparison_and_qualification": "Ge/BSG spheroidization/perturbations versus Ge/ASG; do not fabricate uniform matched settings"
    },
    {
      "id": "C_NECK",
      "branch_ids": [
        "NECK_WITH_CORE",
        "NECK_NO_CORE"
      ],
      "comparison_and_qualification": "Silica neck profiles with/without a core, retaining quench-crack history"
    },
    {
      "id": "C_RAMAN_SI",
      "branch_ids": [
        "RAMAN_RAW_SI",
        "RAMAN_CLAD_SI",
        "RAMAN_REL_SI"
      ],
      "comparison_and_qualification": "Original rod/clad/released; retain polishing state"
    },
    {
      "id": "C_RAMAN_GE",
      "branch_ids": [
        "RAMAN_RAW_GE",
        "RAMAN_CLAD_GE_SILICA",
        "RAMAN_CLAD_GE_ASG",
        "RAMAN_REL_GE"
      ],
      "comparison_and_qualification": "Original Ge/two claddings/released; not pristine stress ground truth"
    },
    {
      "id": "C_CHARACTERIZATION",
      "branch_ids": [
        "CHAR_SI",
        "CHAR_GE"
      ],
      "comparison_and_qualification": "Each retains its original-rod reference, dedicated cross-sections, and lamella child lineage"
    },
    {
      "id": "C_SINGLE_OPTO",
      "branch_ids": [
        "OPTO_SI",
        "OPTO_GE"
      ],
      "comparison_and_qualification": "Si532nm and Ge1550nm; do not arbitrarily change to a common wavelength and call it a source condition"
    },
    {
      "id": "C_DIRECTION",
      "branch_ids": [
        "DIRECTION_SI"
      ],
      "comparison_and_qualification": "Same-object front/side, changing only the relative direction of incidence"
    },
    {
      "id": "C_MECHANICAL",
      "branch_ids": [
        "TENSILE_SI",
        "TENSILE_GE",
        "IMPACT_SI",
        "IMPACT_GE",
        "TORSION_FAIL_SI",
        "TORSION_FAIL_GE"
      ],
      "comparison_and_qualification": "Three metrics for the two single-core devices, with independent destructive sibling samples"
    },
    {
      "id": "C_DURABILITY",
      "branch_ids": [
        "BEND_SI",
        "BEND_GE",
        "COMPRESS_SI",
        "COMPRESS_GE",
        "WASH_SI",
        "WASH_GE",
        "THERMAL_SI",
        "THERMAL_GE"
      ],
      "comparison_and_qualification": "before/after within each episode; symmetric Si/Ge expansion is a task choice and does not imply 8 independent source experiments"
    },
    {
      "id": "C_PPG",
      "branch_ids": [
        "APP_WATCHBAND"
      ],
      "comparison_and_qualification": "Comparison of fibre/commercial proxies under the same phantom preset input, not reproduction of a human experiment"
    }
  ],
  "statistical_pack": {
    "source": "Fig3c",
    "responsivity_NEP_n_per_material": 9,
    "other_Fig3c_metrics_n_per_material": 6,
    "authored_allocation": "Default to independent sibling samples for different metric groups; explicitly cross-reference shared nondestructive groups without assuming the authors' allocation",
    "requirement": "Distinct sample IDs cannot be fabricated from repeated acquisition on the same object; an incomplete statistical package does not establish coverage of the n values"
  },
  "count_warning": "Control packages, cycles, and task configurations cannot be added together as the authors' total experiment count"
}