Show the work.

Published geometry, turns and magnet data support an independent-input EMF comparison. The older section result was 242.8 V (+15%); that historical result used different modeling conditions, so it is not a controlled mesh-convergence series.

*169 N·m is inferred from measured EMF, not an independent shaft-torque measurement. The rated 159 N·m point has an unpublished current angle. Do not use the two torque references interchangeably.

Published turns, conductor and geometry support the comparison. Resistance also tests the modeled lap end turns. Both EMF conventions are shown because the measurement’s RMS convention is not stated.

*The same measured RMS reference is shown for both readings; it is not a separately measured fundamental. Bench magnet temperature is unpublished. Grain-oriented steel is modeled isotropically, so EMF agreement does not establish loss accuracy. The ungraded grid read 12.86 V (12.85 V in the benchmark registry).

The EMF is a closure check: winding turns were fitted to EMF on the section model. Agreement checks consistency of the reconstructed model, rather than independently predicting an unknown measured machine.

*Energies 11(11):2983, Table 3; documented thermal-network result with published support paths and no fitting. Another campaign reports 56.0 °C. Loaded magnet-temperature rises remain 60–85% low in documented cases. Winding-temperature agreement is not a blanket thermal-accuracy claim.

The current–torque curve probes loaded saturation. The loss comparison exposes a remaining discrepancy: measured machined-stator losses are 14–28% above the model; stamped-stator losses are 75–100% above it, across the documented comparison range.

*Measured shaft torque is adjusted by the 1.8 N·m no-load torque to compare electromagnetic torque. †Machined stator; figure readings approximately ±2%. Differences here use (model − reference) / reference; the source’s ‘measurement above model’ uses a different denominator. Slot dimensions and magnet grade are partly assumed. Holder eddy currents and bearing/seal losses are not included.

This reference helped identify end-turn length errors. The current documented model is below the measured terminal resistance; terminal wire is included in the measurement and omitted from the model.

*Derived phase reference from measured coil resistance, not a separate 20 °C phase test. No measured EMF comparator is published. Magnet grade and arc are assumed; comparison against the paper’s analytical EMF is not experimental validation. This winding record is not a newly rerun result.