Why 45% and 81% are not the same MHD efficiency
DMGB | Independent researcher | Evidence-definition note | 5 October 2026 | Working draft

The claim under review
A proposed comparison contrasted a roughly 45 percent kinetic-energy fraction in an Ames accelerator with roughly 81 percent in a cycle study. A difference of 36 percentage points would be meaningful only if both quantities used the same numerator, denominator, flow state and measurement boundary. They do not.

Ames: a core-flow partition
Bogdanoff and Mehta, Experimental Demonstration of MHD Acceleration, AIAA 2003-4285, section IV.E (printed p.19), derive core-flow load factors LF=2.1-2.3 using electrical and velocity diagnostics. Their relation partitions core electromagnetic work into kinetic and Joule contributions. Thus 1/LF gives 43.478-47.619 percent kinetic share and 1-1/LF gives 52.381-56.522 percent Joule share. At LF=2.2 the values are 45.455 and 54.545 percent. This is an author-derived partition, not separate calorimetric measurements of an entire installed accelerator. Sheath and boundary-layer losses are outside this split. The LF range is not a confidence interval.

Schneider: a thermodynamic pressure/enthalpy metric
NASA/TM-2011-217210, printed pp.8-11, reports calculated isentropic efficiencies of 84 percent for the generator and 81 percent for the accelerator. Isentropic efficiency characterizes a thermodynamic process through its pressure and enthalpy relationships. It is not the direct ratio of axial kinetic-energy increase to terminal electrical input. The 81 percent number should not be described as measured, nor simply as an arbitrary fitted replacement for the Ames fraction.

Why the distinction changes the argument
Electrical work can appear as kinetic energy, internal energy and losses at different boundaries. Joule heating is not destroyed energy: some can affect later nozzle expansion, while entropy generation and pressure loss reduce recoverability. An isentropic measure, a core-flow partition, terminal conversion and installed efficiency each answer different questions. Substituting 0.45 for an isentropic 0.81 in an existing cycle relation is not a validated correction. Conversely, using 0.81 as direct terminal-to-kinetic conversion is also unsupported by that source.

Data needed for a legitimate translation
Use common inlet and outlet control surfaces, synchronized mass flux and velocity profiles, total/static pressure and temperature, electrical terminal power, sheath voltage, wall heat, composition and magnetic/cooling auxiliaries. State whether the comparison includes subsequent nozzle expansion. Propagate joint uncertainty and keep a calibration dataset separate from validation cases. A measured change in averaged velocity alone cannot supply these quantities.

Consequence for the threshold map
The map's accelerator axis is an explicitly assumed useful-energy conversion. It is not labeled an Ames whole-device measurement or a Schneider isentropic efficiency. Values 0.45, 0.60 and 0.81 can be inspected as sensitivity values, but only a compatible calibrated dataset could assign them empirical meaning. The useful publication is a definition and evidence audit, not a demonstrated 36-point efficiency deficit.

References
Bogdanoff and Mehta (2003): https://doi.org/10.2514/6.2003-4285
Schneider (2011): https://ntrs.nasa.gov/citations/20110016528
Kato et al. (2004), velocity-response comparison: https://ntrs.nasa.gov/citations/20040105526
