S76 is MetroVolt's integrated systems evaluation: one code, one frozen point, eight coupled constraints that must all pass at once. They do — and the margins are published, including the uncomfortable ones.
The eight gates: Greenwald density fraction (1.16 of limit 1.30), normalized β (4.41 vs 4.50 Troyon), safety factor (q95 4.43 > 2 kink floor in the 0-D convention), peak conductor field (24.6 < 26 T), magnet hoop stress (604 < 800 MPa), confinement requirement (H98 1.89 < 2.0 by the 0-D convention), radial build (1.31 of 1.73 m), and engineering gain (Q_eng 2.78 > 1).
The binding constraint — the one the design leans on hardest — is β/confinement physics, exactly matching the series' own risk statement: MetroVolt is 'a magnet-and-materials programme betting on a confinement result.' The systems code confirms the bet is the right one to interrogate.
Plenty of concepts pass their own chapters; far fewer pass a simultaneous closure test and then publish which wall they stand closest to. Knowing the binding constraint tells investors and researchers precisely which experiment de-risks the most value — that is what makes an honest systems code a capital-allocation tool, not just a physics exercise.
| Constraints evaluated | 8, coupled, simultaneous |
| Result at frozen point | all pass |
| Binding constraint | βN / confinement |
| Engineering gain Q_eng | 2.78 |
| Magnet stress margin | 604 MPa vs 900 yield (1.5×) |