Skip to content

G10 Selected External Campaign

Evidence ID: 026-G10-SELECTED-EXTERNAL-001 Execution source SHA: efed8c3e1bcf173d335b3b9a605febd0fa1084cb Worktree at capture: dirty=false Overall campaign status: PARTIAL

This pack records only the two routes selected by the G10 Owner Review. It does not close G10, reopen G07, promote Total-Lagrangian elasticity, or alter any solver implementation. QF results, external results and derived metrics are kept in separate sections of the machine-readable evidence.

Commands and external tools

  • QF arc-length: run_common_fem_snap_through_benchmark(radius=0.02, max_arc_steps=80)
  • Code_Aster arc-length: scripts/run_code_aster_arc_length_025.py --arc-length-end 1.0 --arc-length-steps 80
  • TET4 TL: scripts/run_code_aster_tl_structural_vnv.py
  • HEX8 TL external: scripts/run_tl_physical_branch_code_aster.py
  • HEX8 QF matched point: one-step QF solve at the first Code_Aster load point
  • Code_Aster: 18.1.0, pinned image recorded in the JSON evidence

All runs used the same clean source SHA efed8c3e1bcf173d335b3b9a605febd0fa1084cb; the result directories are ignored generated output. Input decks and meshes are archived under qualification/0_2_6/g10_selected_external_inputs/.

Arc-length continuation

Classification: PASS_WITH_LIMITATIONS.

Metric QF Solver Code_Aster
Complete path points 80 81
Turning-point load factor -0.0373265417695 -0.0373215949542
Turning-point control displacement -0.893376943327 -0.887500000002
Branch turns 1 1
QF maximum relative residual 2.811e-11 not exported per step

Derived peak differences are 0.013255% in load factor and 0.662191% in control displacement. The common displacement interpolation covers 75 points, with a maximum absolute load-factor difference of 8.868625e-06. The branch and turning point agree qualitatively and quantitatively within the different continuation point placement, but this remains bounded external evidence rather than a new qualification threshold.

Total-Lagrangian TET4

Classification: PASS_WITH_LIMITATIONS.

  • Code_Aster status: PASS_EXTERNAL_CORRELATION.
  • Stress-patch relative error: 8.544217e-05.
  • Imperfect-column maximum relative difference: 1.692986e-09.
  • Complete comparison points: 4 QF / 4 Code_Aster.
  • Formulation: Code_Aster 3D/TETRA4, Green-Lagrange elastic route; same-mesh QF column and stress-patch calculations.

The column path stops at 80 percent of its same-mesh critical load. The external run therefore supports a bounded compatible comparison, not a general finite-deformation claim.

Total-Lagrangian HEX8

Classification: PASS_WITH_LIMITATIONS.

  • Code_Aster status: OBSERVED_EXTERNAL_PATH with 128 exported load points.
  • Matched load factor: 0.0078125.
  • Displacement relative difference: 2.342974e-09.
  • Reaction relative difference after sign alignment: 1.319392e-11.
  • QF matched-point residual: 2.288833e-13; det(F) range 0.999883042065 to 0.999937301077.

The QF full instrumented path was not completed within the bounded campaign budget, so this is a matched first-point comparison only. Code_Aster stress and energy outputs are intentionally not mixed with QF measures.

Decision boundary

This selected campaign remains a PARTIAL intermediate evidence record. The later Owner closeout records G10 as PASS_WITH_LIMITATIONS; no G07, G08, G09, G11 or G12 decision changes are made by this campaign. The selected route records are external evidence with limitations:

  • arc_length_continuation: PASS_WITH_LIMITATIONS;
  • total_lagrangian_elasticity TET4: PASS_WITH_LIMITATIONS;
  • total_lagrangian_elasticity HEX8: PASS_WITH_LIMITATIONS.

Finite-kinematic J2, J2-plus-geometry, coupled contact and triple coupling were not run. They remain at their prior classifications.

See g10_selected_external_evidence.json and g10_selected_external_manifest.json for full compact curves, input digests, runtime provenance, and explicit limitations.