Multi-Span Continuous Beam Solver qualification

Ledger verdict

Verified with errata that must be applied at binding time.

Status

VERIFIED_WITH_ERRATA

Tranche

tranche-8

Review vectors

4

Ledger rows

1

Applied errata

[
  {
    "vector": 2,
    "field": "span_1 reaction split / global A,B",
    "wrong": "left=14.875, right=5.125, A=14.875, B=18.375",
    "correct": "left=5.125, right=14.875, A=5.125, B=28.125 (C=6.75 unaffected)",
    "cause": "span1's left/right reaction split is transposed — confirmed by independent global-moment-about-A check: her triple gives 4*18.375+10*6.75=141, but 4R_B+10R_C must equal 2*20+7*20=180 for the stated loads; the corrected split (28.125,6.75) gives exactly 180. Her own derivation text computes the same 5.125 value and explicitly calls it 'the right-side reaction', then two lines later relabels it 'left' via a sign-flipped M_right/M_left substitution"
  }
]

Stress-test correction

{
  "v1": {
    "vector_inputs": {
      "supports": "A-B-C",
      "span_lengths": [
        4,
        6
      ],
      "loads": [
        [
          {
            "type": "point",
            "magnitude": 10,
            "position_from_left": 1
          }
        ],
        [
          {
            "type": "point",
            "magnitude": 12,
            "position_from_left": 2
          }
        ]
      ],
      "EI": [
        1,
        4
      ]
    },
    "wrong_outputs": {
      "support_moments": {
        "A": 0,
        "B": -7.9090909091,
        "C": 0
      },
      "support_reactions": {
        "A": 0.5227272727,
        "B": 14.7954545455,
        "C": 6.6818181818
      }
    },
    "corrected_outputs": {
      "support_moments": {
        "A": 0,
        "B": -7.0454545455,
        "C": 0
      },
      "support_reactions": {
        "A": 11.1363636364,
        "B": 7.4671717172,
        "C": 3.3964646465
      }
    }
  },
  "v2": {
    "vector_inputs": {
      "supports": "A-B-C",
      "span_lengths": [
        3,
        5
      ],
      "loads": [
        [
          {
            "type": "udl",
            "intensity": 6
          }
        ],
        [
          {
            "type": "point",
            "magnitude": 20,
            "position_from_left": 1
          }
        ]
      ],
      "EI": [
        2,
        8
      ]
    },
    "wrong_outputs": {
      "support_moments": {
        "A": 0,
        "B": -9.1764705882,
        "C": 0
      },
      "support_reactions": {
        "A": 5.9411764706,
        "B": 17.8941176471,
        "C": 14.1647058824
      }
    },
    "corrected_outputs": {
      "support_moments": {
        "A": 0,
        "B": -9,
        "C": 0
      },
      "support_reactions": {
        "A": 12,
        "B": 24.04,
        "C": 1.96
      }
    }
  },
  "method": "independently verified two ways (per-span-EI three-moment theorem via triangular-BMD centroid decomposition; direct-stiffness-matrix FEM solve, 2 elements/3 nodes/fixed-end-moments) — both agree exactly. Reactions recomputed from the corrected moments via standard end-moment reaction correction, and checked against total-load equilibrium (ΣR = ΣP exactly in both vectors)."
}

Adversarial re-test context

The full catalog was re-tested from boundary, convention, and self-check angles in 2,787 requests. Five genuine computational defects were independently verified and corrected. This page reports the current ledger state for this instrument; it does not imply product deployment or interface qualification.

Release gate

Bind every attached correction, replay the affected vectors, enforce boundary validation, and preserve evidence before describing the executable surface as qualified.

Other records

Sources and evidence

These pointers bind the page to evidence. They do not grant authority or collapse verification into deployment.

• VERIFIED_LEDGER.json
• STRESS_TEST_RECEIPT.json
• tranche errata record
• STRESS_TEST_CORRECTIONS_APPLIED.json