Direct Energy Conversion for a D-3He Magnetic-Mirror Burner: the Electron Channel is the Larger Axial Stream, and a Quasineutral Expander Can Recover It — Reproducible Code and Data

Complete manuscript, data, figures, and cold-start reproduction code for a study of direct energy conversion (DEC) in a deuterium–helium-3 magnetic-mirror burner. The conventional framing of advanced-fuel DEC is built on the 14.66 MeV D–³He proton (~80% of charged birth energy); we show this describes the birth spectrum, not the axial loss channel. At an ion temperature of 80 keV the proton thermalizes onto the electrons, so the axial exit carries a thermalized split in which the electron/thermal stream (~55.5%) is larger than the residual directed-ion stream (~44.5%). This relocates the DEC problem onto a directed electron stream, for which no published converter exists (the Richardson–Dushman back-emission wall). We propose a quasineutral, two-species expander DEC — the ambipolar magnetic nozzle run backward — which recovers ~28% of the electron channel in a worked embodiment. It lifts axial electric recovery by ~13% (~97 MWe) but moves the plant energy multiplication Q_E only from 1.0023 to 1.0026, because Q_E is gated by plug/thermal-barrier recirculation, not by DEC. DEC is reported as a bounded upside lever, not the burner's closure mechanism. Every claim is basis-tagged [D] demonstrated, [T] theory, [R] open-risk; the expander is a [T] result with a pre-registered bench experiment (X11→X12) that would convert it to [D]. DESIGNED ≠ MEASURED: Kronos has built no converter; every efficiency is public physics applied to the operating point. CC BY 4.0.