Verification of shearcheck's AISC 360-22 shear tab engine
Every limit state matches two independent blind derivations to within 0.001%, and the printed values of AISC Design Example II.A-17A to within 1%, with the same governing limit state at published precision. This page summarizes the protocol and results; the full verification report (PDF) carries the complete record. The same reference values run as regression tests on every code change.
Protocol
Two independent computational witnesses each derived the complete connection check from a frozen, self-contained prompt restating the cited ANSI/AISC 360-22 provisions and the Brandt (1982) instantaneous-center procedure. Neither witness had access to the engine, to the other witness, or to any worked example. The engine was built separately from the same pinned clause equations. Acceptance thresholds were committed before comparison: engine vs witnesses < 0.5%; engine vs the printed Design Example ±1% with the same governing limit state.
The two witnesses agreed with each other on every quantity to six significant figures — including the ICR coefficient (C = 3.556378) and its converged instantaneous center. The witness derivations, the frozen prompt, and the engine comparison document are archived with the application's verification documents as the provenance record for this page.
Engine vs witness consensus (nominal strengths, kips)
Verified configuration: AISC Design Examples v16.0, Example II.A-17A (four 3/4-in Group 120 N bolts, PL 1/4 × 4 1/2 × 11 1/2 A572 Gr 50, e = a/2 = 1.50 in).
| Quantity | Witnesses | Engine | Δ |
|---|---|---|---|
| Bolt shear (per bolt) | 23.8565 | 23.8565 | 0 |
| Plate bearing (per bolt) | 29.2500 | 29.2500 | 0 |
| Plate tearout, edge | 16.4531 | 16.4531 | 0 |
| Plate tearout, interior | 42.6563 | 42.6562 | ~0 |
| Web bearing | 44.4600 | 44.4600 | 0 |
| Web tearout | 64.8375 | 64.8375 | 0 |
| ICR coefficient C | 3.556378 | 3.556382 | 1×10−4 % |
| Bolt group Rn | 78.2603 | 78.2604 | ~0 |
| Shear yield Rn | 86.2500 | 86.2500 | 0 |
| Shear rupture Rn | 78.0000 | 78.0000 | 0 |
| Block shear Rn | 87.3438 | 87.3438 | 0 |
| Weld Rn | 128.0747 | 128.0747 | 0 |
| Design φRn / Rn/Ω | 58.5000 / 39.0000 | 58.5000 / 39.0000 | 0 |
| Governing limit state | plate shear rupture | plate shear rupture | match |
Engine vs the printed Design Example (available strengths, kips)
Against the values printed in AISC Design Examples v16.0, Example II.A-17A. Differences trace to the example's intermediate rounding, noted per row; the committed tolerance is ±1%.
| Quantity | DE printed | shearcheck | Δ | Note |
|---|---|---|---|---|
| Shear yield φRn | 86.4 | 86.25 | 0.17% | DE rounds Agv to 2.88 in² |
| Shear rupture φRn | 58.5 | 58.5000 | 0.00% | exact |
| Block shear φRn / Rn/Ω | 65.3 / 43.6 | 65.508 / 43.672 | 0.32% / 0.17% | DE rounds areas (1.79 / 0.266) |
| Bolt shear φrn | 17.9 | 17.892 | 0.04% | |
| Plate tearout, edge φrn | 12.4 | 12.340 | 0.49% | DE per-inch table rounding (49.4 k/in) |
| Plate bearing, interior φrn | 22.0 | 21.938 | 0.28% | |
| Web bearing φrn | 33.4 | 33.345 | 0.16% | |
| ICR coefficient C | 3.54 (Table 7-6 interp.) | 3.5564 | 0.46% | exact ICR vs table interpolation |
| Bolt group φRn / Rn/Ω | 58.5 / 38.9 | 58.695 / 39.130 | 0.33% / 0.59% | via exact C |
| Connection φRn / Rn/Ω | 58.5 / 38.9 | 58.5 / 39.0 | 0% / 0.26% |
Governing-limit-state note (disclosed)
The Design Example concludes the connection is controlled by shear transfer at the bolt holes at φRn = 58.5 kips; at the example's printed precision this ties plate shear rupture (also 58.5 kips). With the exact ICR coefficient, the bolt group rises to 58.7 kips and plate shear rupture (58.500 kips) governs. The connection's available strength is identical either way. shearcheck reports the exact-computation governing state.
ICR solver verification
The instantaneous-center-of-rotation solver independently reproduces the two fully-worked coefficients published in Brandt (1982) — Example 1 (C = 1.40) and Example 2 (C = 1.10) — and the Design Example's tabulated coefficient (3.54 interpolated vs 3.5564 exact). See the eccentricity methodology page for the method itself.
Witnessed point vs computed envelope
The independent verification above is at the published example's configuration (n = 4, 3/4-in Group 120 N bolts). The engine's limit-state equations and the ICR solver are fully general within the V1 envelope (n = 2–9, four bolt diameters, two plate grades); the witnessed point and the computed envelope are distinct claims, stated as such. shearcheck output is a calculation aid, not a sealed design — the engineer of record or connection design engineer must review and seal. Not affiliated with AISC.