# Aerospace Evidence Folder - HPT Blade CHT Analysis (COU2)
# Cruise Steady-State Peak Temperature and Creep-Life Prediction
# NASA-STD-7009B Credibility Assessment at MRL 4
# -----------------------------------------------------------
#
# Bundle: aero-evidence-cou2
# Pack:   nasa-7009b
# Decision: NOT ACCEPTED (cruise validation evidence required before MRL 4)
#
# Files (10):
#   cou_definition.docx
#   cfx_solver_settings.txt
#   cascade_reuse_traceability.txt
#   cruise_uq_study.csv
#   sensitivity_study_cruise.csv
#   credibility_assessment_narrative.docx
#   review_board_minutes_cruise_2026Q2.txt
#   risk_assessment_memo.pdf
#   decision_rationale.pdf
#   EVIDENCE_MANIFEST.txt (this file)
#
# Relationship to COU1:
#   COU2 re-uses the SAME CFD model (ANSYS CFX 2025 R1, SST k-omega, CHT
#   coupling, 6.4M cell mesh) and the SAME cascade rig validation data as
#   COU1, applied to a different operating point (cruise vs take-off).
#   All verification evidence transfers cleanly. The gap is entirely in
#   the applicability of the take-off-designed cascade validation to the
#   cruise operating point.
#
# Intentional credibility gaps:
#
#   Five factors recorded as Not Assessed for cruise COU:
#
#     Factor 2.1 Model Form:                 factorStatus = not-assessed
#     Factor 2.4 Test Conditions:            factorStatus = not-assessed
#     Factor 3.1 Equivalency of Inputs:      factorStatus = not-assessed
#     Factor 3.2 Output Comparison:          factorStatus = not-assessed
#     Factor 4.2 Relevance of Validation:    factorStatus = not-assessed
#
#   Root cause: cascade rig designed for take-off Reynolds (1.20e6) does
#   not support assessment at cruise Reynolds (0.85e6). The 41 percent
#   mismatch is regime-changing, not tolerance-bounded.
#
#   These factors are recorded as Not Assessed (rather than assessed at
#   Level 1) because no valid assessment has been performed at cruise
#   conditions. Level 1 would misrepresent the evidence state.
#
# Expected core rule fires (aero-cou2, MRL 4, decision Not Accepted):
#
#   W-EP-04 (Unassessed Factor at Elevated Risk):
#     Fires for each not-assessed factor when modelRiskLevel > 2.
#     MRL = 4, so fires 5 times — once per Not Assessed factor.
#     Severity: High.
#
#   W-AR-02 (Rebutting Defeater — Contradictory Result Ignored):
#     Fires when decision outcome is 'Accepted' AND achievedLevel <
#     requiredLevel. COU2 decision is 'Not Accepted', so W-AR-02
#     DOES NOT FIRE. This is the Morrison-COU2 parity mechanism.
#
# Expected NASA pack rule fires (evidence-linking dependent):
#
#   W-NASA-01 (Data pedigree not-assessed + MRL>1):
#     Does NOT fire. Data Pedigree is assessed at Level 3.
#
#   W-NASA-02 (Dev technical review assessed + no evidence linked):
#     Fires if import does not link review_board_minutes_cruise_2026Q2.txt
#     as ReviewActivity evidence on the tech-review factor.
#
#   W-NASA-03 (Process mgmt assessed + no evidence linked):
#     Fires if import does not link configuration management evidence.
#
#   W-NASA-04 (Use history not-assessed + MRL>2):
#     Does NOT fire. Use History is assessed (Level 1 against Level 2).
#
#   W-NASA-05 (Results uncertainty assessed + no UQ linked):
#     Fires if import does not link cruise_uq_study.csv as
#     uofa:hasUncertaintyQuantification evidence.
#
#   W-NASA-06 (Results robustness assessed + no sensitivity linked):
#     Fires if import does not link sensitivity_study_cruise.csv as
#     SensitivityAnalysis evidence.
#
# Structural parallel to Morrison COU2:
#   Morrison COU2: 6 weakeners, W-EP-04 only (6 not-assessed factors at
#                  MRL 5), decision Not Accepted, zero W-AR-02.
#   Aero COU2:     5 weakeners from W-EP-04 (5 not-assessed factors at
#                  MRL 4), decision Not Accepted, zero W-AR-02. Plus
#                  possible NASA pack fires from evidence-linking
#                  variability and core structural fires.
#
#   Core mechanism is identical: focused single-pattern firing on
#   not-assessed factors at elevated risk; decision Not Accepted keeps
#   W-AR-02 at zero.
#
# Narrative thesis for NAFEMS:
#   "Same tool, same rules, same model, same evidence package — but a
#    different COU produces a structurally different credibility outcome.
#    COU1 (take-off, Accepted) fires W-AR-02 multiple times and
#    COMPOUND-01. COU2 (cruise, Not Accepted) fires W-EP-04 five times
#    with zero W-AR-02. The divergence is entirely a function of
#    evidence-COU fit, not tool behavior. The machine detects this at
#    import time; the human would need to read 19 pages to reach the
#    same conclusion."
#
# USAGE:
#   uofa extract aero-evidence-cou2/ --pack nasa-7009b -o aero-cou2.xlsx
#   # human review of xlsx - confirm 5 factors marked Not Assessed
#   uofa import aero-cou2.xlsx -o aero-cou2.jsonld
#   uofa check aero-cou2.jsonld
#   uofa reason aero-cou2.jsonld -o aero-cou2-reasoned.jsonld
#   # expect:
#   #   - 5x W-EP-04 fires (on factors 2.1, 2.4, 3.1, 3.2, 4.2)
#   #   - 0x W-AR-02 fires (decision is Not Accepted)
#   #   - possible NASA pack fires from evidence-linking
