feat: add assembly reduced solver boundary
This commit is contained in:
+29
-1
@@ -13,10 +13,38 @@ Every new agent session must read this file together with `PLAN.md` before plann
|
||||
- Do not remove history unless the user explicitly asks for archival cleanup.
|
||||
|
||||
## Current Status
|
||||
Phase 1 has a new rebaseline phase definition in `phases/1-linear-static-mitc4-rebaseline`. Steps 0 through 11 are complete. `quad_02_phase1.inp` is now the normalized Phase 1-compatible input path for the stored `quad_02` S4 reference pair, while the original `quad_02.inp` remains preserved unsupported provenance. Core numeric aliases, DOF mapping, validation harness, model diagnostic context, the Phase 1 parser/domain subset, validation/singular diagnostics, DofManager/reaction foundation, minimum result model metadata, displacement CSV comparator foundation, MITC4 geometry/director scaffolding, MITC4 displacement/strain/tying row scaffolding, MITC4 material/transform/integration scaffolding, MITC4 stiffness/drilling/internal-force scaffolding, and MITC4 patch/locking-sensitivity tests have been revalidated. The old `phases/1-linear-static-mitc4` path is historical and superseded after the MITC4 formulation reset.
|
||||
Phase 1 has a new rebaseline phase definition in `phases/1-linear-static-mitc4-rebaseline`. Steps 0 through 12 are complete. `quad_02_phase1.inp` is now the normalized Phase 1-compatible input path for the stored `quad_02` S4 reference pair, while the original `quad_02.inp` remains preserved unsupported provenance. Core numeric aliases, DOF mapping, validation harness, model diagnostic context, the Phase 1 parser/domain subset, validation/singular diagnostics, DofManager/reaction foundation, minimum result model metadata, displacement CSV comparator foundation, MITC4 geometry/director scaffolding, MITC4 displacement/strain/tying row scaffolding, MITC4 material/transform/integration scaffolding, MITC4 stiffness/drilling/internal-force scaffolding, MITC4 patch/locking-sensitivity tests, full-space assembly, reduced projection, sparse-pattern scaffold, solver adapter injection, and full-vector internal/reaction force state have been revalidated. The old `phases/1-linear-static-mitc4` path is historical and superseded after the MITC4 formulation reset.
|
||||
|
||||
## Completed Work
|
||||
|
||||
### 2026-05-04 - P1R-12 assembly sparse solver path completed
|
||||
Author: Codex
|
||||
|
||||
Changed files:
|
||||
- `include/fesa/fesa.hpp`
|
||||
- `tests/test_main.cpp`
|
||||
- `phases/1-linear-static-mitc4-rebaseline/index.json`
|
||||
- `PLAN.md`
|
||||
- `PROGRESS.md`
|
||||
|
||||
Summary:
|
||||
- Added deterministic reduced sparse-pattern scaffolding from `DofManager` element equation connectivity, preserving int64 equation and nonzero counts for the future sparse/MKL path.
|
||||
- Added `projectToReducedSystem()` so constrained/free projection is a named assembly boundary instead of being embedded directly inside `LinearStaticAnalysis`.
|
||||
- Extended `AssemblyResult` to carry full-space stiffness, full external load, reduced sparse pattern, and diagnostics.
|
||||
- Reworked assembly to call the rebuilt MITC4 stiffness result path directly and preserve full-space `K_full`/`F_full` for reaction recovery.
|
||||
- Added solver injection to `LinearStaticAnalysis` so the deterministic Gaussian solver remains the default test adapter while future MKL-backed solvers stay behind the `LinearSolver` interface.
|
||||
- Added `AnalysisState::f_internal_full` and verified `RF = Fint_full - Fext_full = K_full * U_full - F_full`.
|
||||
- Added tests for deterministic sparse pattern ordering, reduced projection with known displacement, full-space assembly/load preservation, reduced residual satisfaction, solver adapter injection, and singular solver diagnostic propagation.
|
||||
|
||||
Verification:
|
||||
- First ran `python scripts/validate_workspace.py` after adding Step 12 tests; it failed as expected because reduced sparse pattern, projection, solver injection, and `f_internal_full` did not exist yet.
|
||||
- After implementation, `python scripts/validate_workspace.py` configured CMake, built `fesa_core` and `fesa_tests`, and ran CTest successfully.
|
||||
- CTest result: 1 test executable passed.
|
||||
|
||||
Follow-up:
|
||||
- Continue with P1R-13 linear static workflow revalidation from Phase 1 input through `U` and `RF` result fields.
|
||||
- Step 12 intentionally did not add MKL/TBB APIs, stored Abaqus reference comparison, pressure loads, or a production sparse matrix storage backend.
|
||||
|
||||
### 2026-05-04 - P1R-11 MITC4 patch and benchmark tests completed
|
||||
Author: Codex
|
||||
|
||||
|
||||
Reference in New Issue
Block a user