Files
FemLite/docs/v6_acceptance_report.md
T
shenw ee18a9b757
FemLite CI / build-test (push) Has been cancelled
Release FemLite 0.6.2 stability hardening
2026-07-02 08:26:46 +08:00

6.6 KiB

FemLite 0.6.x Acceptance Report

Date: 2026-07-01

Status: release close-out candidate.

Scope

FemLite 0.6.0 closes the first algorithm-foundation loop after the 0.5.x V5 diagnostic hardening line. The release adds a shared nonlinear lifecycle module, lifecycle hooks for the covered static/Newmark/Riks paths, rollback audit adapters, lifecycle CSV output fields and restart version 5 lifecycle diagnostics.

0.6.0 does not claim delivery of complete industrial AL/mortar consistent tangents, generalized mixed-control continuation, cohesive contact mainline, DEM/FEM subcycling, nonzero harmonic cyclic solving or production roll-forming workflow support.

Delivered Evidence

Item Evidence
Lifecycle ids and snapshot nonlinear_lifecycle covers event names, empty snapshot, history and audit names.
Static nonlinear hook solver_implicit covers accepted and rejected/cutback lifecycle paths.
Nonlinear Newmark hook solver_implicit covers nonlinear Newmark success and rejection.
Riks/arc hook solver_implicit covers lifecycle accept, load factor, arc residual and path append.
Rollback audit Material, contact, cohesive and particle audit adapters feed the shared lifecycle status.
Output schema continuation_path.csv appends lifecycle fields without reordering older columns.
Restart schema Restart version 5 writes optional LIFECYCLE_DIAGNOSTICS; reader accepts files without the block.
Documentation Design, algorithm, interface, test, user and developer docs state lifecycle boundaries and deferred solver rewrites.

Planned Close-Out Commands

cmake --build build-codex-gcc --parallel
ctest --test-dir build-codex-gcc -R "^(solver_implicit|implicit_inp|restart|contact_mortar_report|contact_multiplier_report|cohesive_contact|particle_acceptance)$" --output-on-failure
ctest --test-dir build-codex-gcc -L v5_acceptance --output-on-failure
git diff --check

Results

Command Result
cmake --build build-codex-gcc --parallel passed; CMake regenerated and full build completed.
0.6.0 targeted regression command 7/7 passed.
ctest --test-dir build-codex-gcc -L v5_acceptance --output-on-failure 39/39 passed after updating version assertions to 0.6.0.
git diff --check passed; Git reported line-ending normalization warnings only.

Close-Out Conclusion

FemLite 0.6.0 meets the planned lifecycle foundation acceptance criteria on the local build-codex-gcc configuration. The release remains bounded to diagnostic lifecycle infrastructure and rollback audit coverage for the documented subset.

Deferred Work

  • Complete industrial AL/mortar consistent tangent and full contact state-machine rewrite.
  • Generalized mixed-control continuation and branch switching.
  • Complete shell preload K_G and high-order solid geometric stiffness.
  • Cohesive contact mainline integration.
  • Nonzero harmonic cyclic symmetry.
  • DEM/FEM subcycling and large-particle long runs.
  • Production roll-forming process chain.
  • Additional solver scaling and block preconditioning work.

0.6.1 Diagnostic Hardening Addendum

0.6.1 keeps the 0.6.0 release boundary and hardens diagnostics around the covered nonlinear/contact paths. It adds lifecycle summary queries, cutback-reason taxonomy, nonlinear_lifecycle.csv, stable contact state counters, compact contact long-path checker coverage, AL/friction summary fields, a shared tangent-audit report shell, Riks/arc path indicators, restart-after-each-accepted audit coverage and fuller linear-solver fields in nonlinear rows.

Final 0.6.1 evidence on build-codex-gcc:

cmake --build build-codex-gcc --parallel
ctest --test-dir build-codex-gcc -R "^(nonlinear_lifecycle|solver_implicit|implicit_inp|restart|contact|contact_mortar_report|contact_multiplier_report|cohesive_contact|particle_acceptance|linear_solver_cg)$" --output-on-failure
ctest --test-dir build-codex-gcc -R "^example_(contact_two_blocks_compression|contact_block_sliding|shell_shell_contact|shell_solid_contact)_benchmark$" --output-on-failure
ctest --test-dir build-codex-gcc -L v5_acceptance --output-on-failure
git diff --check

0.6.1 targeted regression passed 10/10, the compact contact benchmark smoke passed 4/4, the V5 acceptance subset passed 39/39 and the version-specific smoke checks passed 2/2 after the public version was raised to 0.6.1.

0.6.1 still does not claim complete industrial AL/mortar consistent tangent delivery, generalized mixed-control continuation, cohesive contact mainline, DEM/FEM subcycling or production roll-forming workflow support.

0.6.2 Stability Close-Out Addendum

0.6.2 is a stability hardening release. It does not add user-facing algorithms or broaden the 0.6.1 feature boundary. The release stabilizes the existing diagnostic surface by hardening CSV schemas, path storage reset/capacity, restart compatibility, contact state defaults, OpenMP contact determinism, linear-solver fallback diagnostics and INP/CLI error boundaries.

Required 0.6.2 gate on build-codex-gcc:

cmake --build build-codex-gcc --parallel
ctest --test-dir build-codex-gcc -R "^source_size$" --output-on-failure
ctest --test-dir build-codex-gcc -R "^(nonlinear_lifecycle|solver_implicit|implicit_inp|restart|contact|contact_mortar_report|contact_multiplier_report|cohesive_contact|particle_acceptance|linear_solver_cg|c_api|cli_output_dir_boundary|vtk_output)$" --output-on-failure
ctest --test-dir build-codex-gcc -L v5_acceptance --output-on-failure
git diff --check

The source_size command is a mandatory 0.6.2 release blocker. It would have caught the hard failures in nlc_path_append, output_csv_write_path and the restart contact-history test before release.

0.6.2 remains bounded to release hardening: CSV schema stability, restart compatibility, deterministic contact diagnostics, solver fallback reporting, INP/CLI error boundaries and documentation/version consistency.

Final local 0.6.2 evidence on build-codex-gcc:

  • cmake --build build-codex-gcc --parallel passed.
  • ctest --test-dir build-codex-gcc -R "^source_size$" --output-on-failure passed 1/1.
  • Targeted stability regression passed 13/13.
  • ctest --test-dir build-codex-gcc -L v5_acceptance --output-on-failure passed 39/39.
  • Version smoke checks passed 2/2 after raising the public version to 0.6.2.
  • git diff --check passed with line-ending normalization warnings only.

0.6.2 explicitly defers complete industrial AL/mortar consistent tangent delivery, generalized mixed-control continuation, cohesive contact mainline, DEM/FEM subcycling, complete shell/high-order geometric stiffness expansion, nonzero harmonic cyclic solving and production roll-forming workflows.