Per the 2026-09-07 checkpoint, reconciliation is narrowed to every completed run under the frozen route: the alpha-Fe positive control (action 01a07de3-7da0, route Ms 2.2967 T). No secondary-arm CHGNet runs from the revised item 6 exist yet.
Inputs (from validated cell and returned per-site moments):
Per-site moment (single Fe 8c-equivalent site): 2.337 uB
Validated lattice constant: a = 2.873 A (bcc, Im-3m, 1 atom/primitive cell)
Primitive cell volume: V = a^3/2 = 11.858 A^3 = 11.858e-30 m^3
Recomputation:
Total moment per formula unit: M = 2.337 uB = 2.337 x 9.2740e-24 A m^2 = 2.1673e-23 A m^2
Magnetization: Ms = M/V = 2.1673e-23 / 11.858e-30 = 1.8277e6 A/m
Flux density: mu_0 Ms = 4pi x 1e-7 x 1.8277e6 = 2.2965 T
Quantity | Route Summary | Independent Recompute | Delta | Flag (>5%) |
|---|---|---|---|---|
Total moment/f.u. (uB) | 2.337 | 2.337 | 0.00% | no |
Cell volume (A^3/f.u.) | 11.858 | 11.858 | 0.00% | no |
Ms (mu_0 T) | 2.2967 | 2.2965 | 0.01% | no |
No unit, multiplicity, or volume disagreement larger than 5%. The route summary for the alpha-Fe control is internally consistent with the returned per-site moments and validated cell volume. The +6.7% signed error against the 2.152 T room-temperature experimental anchor is a model bias, not a bookkeeping disagreement, and remains the only signed error in the panel.
Source action: 01a07de3-7da0 (alpha-Fe positive control, frozen v1 route)
Experimental anchor: 2.152 T, room-temperature polycrystalline alpha-Fe (Moss reference)
This reconciliation adds no new computations; it audits the existing route output only