Files
FemLite/docs/v5_benchmark_matrix.md
T
shenw 75d4e735b1
FemLite CI / build-test (push) Has been cancelled
Release FemLite 0.5.2 V5 closeout
2026-07-01 14:51:40 +08:00

16 KiB
Raw Blame History

FemLite V5 Benchmark And Acceptance Matrix

日期:2026-06-26

本文档记录当前 V5 已完成能力对应的测试入口、示例目录、参考指标和容差。 状态标记仍只维护在 docs/v5_development_plan.md;本文件只给出可执行验收矩阵。

2026-07-01 V5.12 Close-Out Note

This matrix records executable V5 acceptance entries for the bounded delivery subset. It does not treat deferred high-fidelity production workflows as V5 acceptance failures.

Included in the V5 close-out subset: selected two_arches reference-point coverage, bounded spherical-cap continuation/Riks evidence, compact Mohr-Coulomb strength-reduction diagnostics, Euler and compact corrugated buckling checks, small dense generic *BUCKLING, static-preload-to-buckling for explicit line and C3D4/C3D8 solid initial-stress K_G, WAVE=0 explicit-pair cyclic symmetry for solid/shell nodes, compact roll-forming and cohesive benchmarks, and source-size governance regression.

Excluded from V5 acceptance and deferred after V5: full shell preload-buckling geometric stiffness, higher-order solid geometric stiffness, full shell FE corrugated preload buckling, automatic cyclic pairing, nonzero harmonic-wave cyclic semantics, production roll-forming workflow, production cohesive contact workflow, high-fidelity example upgrades dependent on those algorithms, and broad longrun/performance matrices.

Current local V5.12 targeted regression evidence on build-codex-gcc: the bounded buckling/cyclic subset command with implicit_inp, cyclic_symmetry, cli_buckling_generic_dense, cli_buckling_static_preload_dense, cli_buckling_solid_preload_dense, buckling_support, example_euler_column_buckling_benchmark, and example_buckling_corrugated_shell_benchmark passed 8/8.

Current local V5 acceptance subset evidence on build-codex-gcc: ctest --test-dir build-codex-gcc -L v5_acceptance --output-on-failure passed 39/39.

测试层级

层级 用途 推荐命令
默认回归 快速确认全部单元、模块、CLI、示例和 benchmark 不回退 ctest --test-dir build-win-gcc-16 -C Release --output-on-failure
V5 验收子集 只运行当前 V5 已完成能力的核心测试和 benchmark ctest --test-dir build-win-gcc-16 -C Release -L v5_acceptance --output-on-failure
示例 benchmark 只运行带 benchmark 标签的示例级检查 ctest --test-dir build-win-gcc-16 -C Release -L benchmark --output-on-failure
定向调试 调试单个能力或相邻能力 ctest --test-dir build-win-gcc-16 -C Release -R "<test-name>" --output-on-failure

默认回归当前 Windows GCC 16 Release 记录为 106/106 passed。V5 验收子集 当前记录为 35/35 passed。

多平台复测矩阵

平台 构建目录 工具链 默认 CTest V5 验收子集 备注
Windows GCC build-win-gcc-16 GNU Fortran 16.1.0 106/106 passed 35/35 passed 20.1 clean clone 和 20.2 后本地回归均通过
WSL GNU build-wsl-gfortran-v5-20-2 GNU Fortran 13.3.0、GNU C 13.3.0、CMake 3.28.3 106/106 passed 35/35 passed benchmark 工具已修正跨平台路径分隔符
Windows Intel/MSVC build-win-ifx-v5-20-2 VS DevCmd 17.14.33、MSVC 19.44.35227.0、IntelLLVM ifx 2026.0.0 106/106 passed 35/35 passed CTest 环境优先加载 ifx compiler runtime;Mohr-Coulomb 边界分类容差已稳定
Windows GCC OpenMP build-win-gcc-16 GNU Fortran 16.1.0,OpenMP enabled contact_omp1、contact_omp4、perf_contact_pair_omp_consistency 3/3 passed n/a 1/4 线程 contact/performance smoke 非计时统计一致;perf_contact_pair_omp_consistency 当前覆盖 64 面片 4 步 force/tangent/cache 路径

V5 验收矩阵

V5 项 能力 目录或模块 CTest 名称 参考指标 容差/验收条件
V5.2 内部 solver 诊断、no-external-solver 策略 src/solver, tests/module, tests/interface linear_solver_cg, csr_diagnostics, solver_implicit, implicit_inp, c_api, CLI implicit/static tests CG/BiCGSTAB/GMRES/dense pivoting、残差、CSR matrix class、legacy external id rejection、非线性修正 fallback 旧外部 method/backend id 被拒绝且不改变设置;静力、隐式和非线性路径通过内部 solver dispatch;非线性修正可在内部求解器间 fallback 且恢复用户设置
V5.3 非匹配实体 mortar 压缩 examples/contact_two_blocks_compression example_contact_two_blocks_compression_benchmark pressure_profile.csv, active contact, master coverage, weighted gap, multiplier residual 多 master facet 覆盖;weighted penetration 和 multiplier residual 在 checker 容差内;VTK 输出存在
V5.3 大滑移 mortar-friction examples/contact_block_sliding example_contact_block_sliding_benchmark sliding_profile.csv, stick/slip, accumulated slip, tangential force, friction work 多 master/active candidate;累计滑移和摩擦功为正且非负耗散;VTK 输出存在
V5.3 shell-shell contact examples/shell_shell_contact example_shell_shell_contact_benchmark contact profile, active contact, force norm, gap residual, augmented multiplier active contact 大于零;gap/force/multiplier 指标通过 checker
V5.3 shell-solid contact examples/shell_solid_contact example_shell_solid_contact_benchmark contact profile, master search, shell-to-solid force transfer active contact 大于零;master search 和 nodal force 指标通过 checker
V5.3 接触状态、restart 和并行路径连续性 tests/module, tests/performance contact_mortar_al, contact_mortar_iter, contact_mortar_report, contact_mortar_tangent, contact_multiplier_report, restart, perf_contact_pair_omp_consistency AL multiplier, contact_multiplier_audit, mortar_tan_audit_matrix, tangent/report diagnostics, restart restored history, 64 面片 4 步 OpenMP path stats 接触历史、摩擦累计量、stick/slip 状态计数和 multiplier 状态在 restart 后连续;0.5.1 覆盖 active/released/open multiplier 写读、duplicate/invalid 为零、commit/trial 差异为零,并检查 mortar tangent row/column balance、刚体平移残差、非对称量和能量 probe;OpenMP 1/4 线程下 force/tangent/cache 路径非计时统计一致
V5.4 Mohr-Coulomb 材料点 tests/unit material_mohr_coulomb, material_mc_driver, material_mc_envelope 三轴压缩/伸展、纯剪、张拉截断、branch id、plastic multiplier yield margin、分支号、塑性增量和 envelope safety ratio 通过测试容差
V5.4 Mohr-Coulomb 屈服面算例 examples/mohr_coulomb_yield_surface example_mc_yield_surface_benchmark yield_surface.csv, yield_points.csv, yield_summary.csv; 原始材料和 F=2 折减材料 压缩包络、伸展包络和张拉截断符合解析公式;强度折减降低承载力;elastic/shear/tension/mixed 分支全部被代表性应力点激活
V5.4 Mohr-Coulomb 边坡强度折减 benchmark examples/strengthReduction example_strength_reduction_benchmark 8 个参考工况,最大 FOS 误差 6.9392E-03,86 trials,52 stable,34 unstable;--nonlinear-chain 可选输出真实 FE trial、cutback、accepted-time 和线性求解诊断;--nl-write-cloud 可选输出目标 FOS 云图 summary 与 VTK;0.75 m cloud route 支持内部 BiCGSTAB/ILU0、静力 predictor、自适应跳步和 accepted segment 后 VTK 刷新 0.75 m 网格和参考几何记录;输入使用原生实体 DLOAD, GRAV;上下土体界面必须为共享节点 conforming mesh;FOS 曲线在 0.02 容差内;FOS 随坡角单调降低;关联工况不弱于 Davis B 工况;plastic zone 和 displacement probe 激活;optional nonlinear chain 不伪造 VTK;optional partial/target cloud VTK 必须含 displacement/eqps 字段且网格计数匹配;0.75 m 临界失稳位移/eqps 云图属于后续 longrun/研究验证,不作为 V5 交付验收项
V5.5 Euler column 解析屈曲 examples/euler_column_buckling example_euler_column_buckling_benchmark first load relative error 3.27660856E-05, residual 1.50014805E-11, phi^T K_G phi=1 相对误差、残差和几何刚度归一化通过 checker
V5.5 波纹壳循环屈曲 benchmark examples/buckling_corrugated_shell example_buckling_corrugated_shell_benchmark selected wave 4 load 453.031407, critical wave 8 load 393.369677, cyclic/full relative error 1.25473461E-16 循环边界残差、模态残差、wave sweep 和 modal CSV 通过 checker
V5.5 屈曲/循环模块 tests/module; CLI buckling_support, cyclic_symmetry, implicit_inp, cli_buckling_generic_dense, cli_buckling_static_preload_dense, cli_buckling_solid_preload_dense geometric stiffness helper, cyclic map residual, small dense generic *BUCKLING solve, WAVE=0 cyclic MPC expansion, C3D4/C3D8 solid initial-stress KG 内部 dense 特征值、Euler helper、周期映射、通用 buckling 小规模 dense 路径、预载派生线 KG、实体 initial-stress KG 和实体/壳 cyclic 映射通过模块/CLI 容差
V5.6 辊运动学和正则化牵引 examples/roll_forming example_roll_forming_kinematics_benchmark, rigid_roller mean slip 0.4 with spin vs 0.2 without spin; mean traction 289.208274 vs 228.478247; torque 14.4604137 有/无角速度路径必须产生不同 slip/traction;摩擦能和扭矩为正
V5.6 塑性成形和回弹 benchmark examples/roll_forming example_roll_forming_springback_benchmark springback ratio 0.473154068, residual displacement 0.0316107559, plastic work 1.11591736E-05, 56/81 plastic points 塑性功为正;塑性区激活;残余弯矩接近零;回弹比例在 checker 范围内
V5.7 cohesive 单积分点 tests/module cohesive_contact bonded loading, mixed-mode damage, unloading/reload, complete decohesion, restart 损伤不可逆;耗散非负;failed 状态 traction 清零;restart 保持状态
V5.7 球-膜粘附/脱粘 benchmark examples/ball_adhesion_decohesion example_ball_adhesion_cohesive_benchmark peak pull-off 119.246023, min compression -362.051100, dissipated energy 0.451891017, final failed fraction 1.0 滞回面积、耗散能、损伤和脱粘面积通过 checker
V5.8 颗粒基础验收 tests/module particle_acceptance, particle_fem_coupling free fall, wall collision, three-particle pile, particle-facet force balance 质量守恒;重叠有界;耗散非负;facet 节点力平衡残差在容差内
V5.8 动态颗粒柔性容器 benchmark examples/Dynamic_Grain_Packing_of_a_Flexible_Container example_dynamic_grain_container_benchmark 4 particles, 4 facet contacts, mapped_steps=40, max load 3.68E+01, min center uz=-7.29240354E-06, max overlap 3.00615680E-03 结构输入不得含 *DLOAD pressure surrogate;映射节点力和容器位移非零;质量守恒、重叠和能量通过 checker
V5.11 finite-strain S4 bounded spherical-cap postbuckling examples/spherical_cap_with_central_point_load default: example_spherical_cap_point_load_benchmark; extended: FEMLITE_SPHERICAL_CAP_NONLINEAR=1 python examples/spherical_cap_with_central_point_load/check_spherical_cap_results.py ... Shell-16 baseline geometry and samples; Shell-4 all-S4 finite-strain target with translational rim support; bounded nonlinear run to uz=-3.24 m, 54 accepted increments, final wn=15.89482, final Fn=1.058395, sampled Fn=1.340817 at wn=13.642 Default CTest remains fast and checks baseline/target口径. Extended nonlinear command validates continuation-path columns, production finite-strain S4 dispatch and a post-peak load drop. Full uz=-6.2 m PDF curve matching, Shell-4/8/16 digitized comparison, pseudo-arclength and late GMRES optimization are deferred.
V5.12 method-4 load-control Riks post-limit branch evidence examples/spherical_cap_with_central_point_load optional: FEMLITE_SPHERICAL_CAP_RIKS=1 python examples/spherical_cap_with_central_point_load/check_spherical_cap_results.py ... Shell-4 finite-strain target converted to proportional crown CLOAD; 80 accepted increments; peak_lambda=0.6909629, final_lambda=0.6017746, final displacement magnitude 1.3973472 m Validates the production load-type Riks augmented equation and predictor direction memory on a real shell structure limit point and post-limit branch segment. Broader branch-following generalization on a second independent structure remains deferred.

示例输出文件

示例目录 主要输出 checker
examples/contact_two_blocks_compression pressure_profile.csv, VTK check_contact_results.py
examples/contact_block_sliding sliding_profile.csv, VTK check_contact_results.py
examples/shell_shell_contact shell contact profile, VTK check_contact_results.py
examples/shell_solid_contact shell-solid contact profile, VTK check_contact_results.py
examples/mohr_coulomb_yield_surface yield_surface.csv, yield_points.csv, yield_summary.csv check_mc_yield_surface_results.py
examples/strengthReduction strength_reduction_path.csv, strength_reduction_cases.csv, summary, plastic_zone.csv, displacement_probe.csv; optional strength_reduction_nonlinear_chain.csv and nonlinear summary when --nonlinear-chain is used; optional strength_reduction_cloud_summary.csv, progress/trace CSV and refreshed partial/target VTK when --nl-write-cloud is used check_strength_reduction_results.py
examples/euler_column_buckling buckling_summary.csv, buckling_mode_1.csv check_buckling_results.py
examples/buckling_corrugated_shell summary, wave sweep, modes, profile check_buckling_results.py
examples/roll_forming roller kinematics history/summary, springback history/summary, plastic_zone.csv check_roller_kinematics.py, check_springback_results.py
examples/ball_adhesion_decohesion adhesion_history.csv, adhesion_summary.csv, decohesion_profile.csv check_cohesive_results.py
examples/Dynamic_Grain_Packing_of_a_Flexible_Container grain_container_history.csv, grain_container_summary.csv check_grain_container_results.py
examples/two_arches nodes/elements/contact/history CSV from the two-arches geometry/contact smoke; optional single displacement reference-point run writes two_arches_reference_point.csv with para, disp_upper, reference mode, contact status and contact force smoke: check_two_arches_results.py; optional reference point: run_two_arches_reference_point.py and check_two_arches_reference_point.py
examples/spherical_cap_with_central_point_load continuation_path.csv, spherical_cap_load_displacement.csv check_spherical_cap_results.py
examples/spherical_cap_with_central_point_load optional Riks continuation_path.csv, spherical_cap_riks_load_path.csv check_spherical_cap_results.py with FEMLITE_SPHERICAL_CAP_RIKS=1

Smoke 与 benchmark 边界

  • V4 smoke 仍用于快速确认旧路径,例如动态颗粒容器目录中的 pressure/ground smoke、roll forming friction smoke 和 ball adhesion contact smoke。
  • V5 完成项必须至少有模块级验收或示例级 checker;不能只依赖 smoke。
  • benchmark 编排工具放在 tools/benchmarks,示例输入和 checker 放在对应 examples/<case>/ 目录,生产功能代码仍保留在 src。
  • 当前模块不引入第三方 solver 或外部求解器进程;相关验收只覆盖内部 solver。

文档治理边界

  • 状态标记只在 docs/v5_development_plan.md 维护;本矩阵只记录可运行入口、 指标和容差,避免多个文档出现互相冲突的 PASS/PARTIAL 口径。
  • 每次新增或修正 benchmark 时,必须同步本文件中的 CTest 名称、示例目录、 checker、参考指标和容差说明。
  • 每次完成发布确认、多平台复测或 longrun 后,必须同步本文件的测试层级和 多平台复测矩阵,并同步 docs/v5_acceptance_report.md 的验收结论。
  • 若某项仅为 smoke、surrogate 或 debug fallback,不得在本矩阵中列为 V5 完成项;只能在边界说明或 DEFERRED/PARTIAL 说明中登记。