Saved heat transport implies inward flux at the rejected MAST-U face
We reviewed the completed reference diagnostic without running another physics solver. Its raw negative heat diffusivities already explain why the guard uses fallback. The next question is whether their signs arise when heat flux is converted to diffusivity.
We reconstructed temperatures and geometry from the saved trajectory and inverted the recorded conversion. All negative channels have temperatures decreasing outward. The conversion’s denominator correction changes none of their signs. The resulting fluxes are algebraically inferred values, not independently saved solver outputs. Their magnitudes are below the floor used by the offline flux accuracy gate, which supplies no guarantee in this region.
| evidence | result |
|---|---|
| signed heat conversion | the final electron heat flux is inferred as -0.001689376033 gB; all 9 negative channel records at 8 accepted states have outward-decreasing temperature and no denominator-epsilon sign reversal, with maximum epsilon ratio 1.03e-11; all negative magnitudes are below 1 gBC-195 |
| conversion and software checks | 33 focused checks pass in each mode, including 14 new sign/units/history tests; all 80 native heat-converter round trips meet the mixed 1e-12 parity criterion (maximum absolute difference 8.88e-16 m²/s), 32 numerical source files match the diagnostic bindings, and all 78 source modules pass typingC-196 |
| formal candidate temperature comparison | remains not evaluable |
The guard checks raw transport before the diffusivity floor is applied. Moving clipping before that check would change the reference model. The current nonnegative surrogate heat outputs also cannot reproduce inward heat flux. Neither observation establishes a physically justified replacement closure.
The installed in-memory solver wrapper does not expose mode spectra. Another unchanged trajectory replay would therefore leave the mode evidence missing. A richer local-state diagnostic is proposed; its output route and compute request remain to be prepared and reviewed. No new solver, training or cluster job ran. The current model, guard, faces and tolerances remain selected; K-030 stays open. Nothing was synchronized, replaced or published.
What’s next. Owner review of reference physics, followed by preparation of any additional diagnostic or changed closure under its own concrete review.
Technical details → Verification › signed heat conversion review
Decisions → OQ-32