The Aegis Forward-Base Burner: Q_E Closure, Plug Requirement, and Mirror Confinement in an Islanded Low-Neutron D–³He Tandem Mirror

Aegis is the forward-base configuration of the Kronos deuterium–helium-3 tandem-mirror burner: a fixed, hardened, grid-independent installation that turns an aneutronic-dominated fusion source into firm electrical power for a defence site without a fuel-convoy tail. This paper is a comprehensive, review-article treatment of the integrated physics case for that machine, and it separates, throughout, what is robust from what is conditional. At the frozen design point the plant reaches an engineering gain =1.318 at fusion power 4298.5 MW, of which 3868.0 MW is the charged-particle D+ ^3He p (14.68 MeV)+α channel and only 233.95 MW is neutron power — a neutron fraction f_n=5.44 %, the low-activation advantage that motivates the fuel and the siting. The binding physics gate is the end-plug potential: confining the central-cell ions requires a plug electrostatic potential =248.75 keV, which through the ambipolar Boltzmann relation demands a plug-to-central density ratio n_p/n_c=(e/)=16 at carrier temperature =89.7 keV — a requirement more than two orders of magnitude beyond the demonstrated tandem-mirror record, retired only by a dedicated plug-potential experiment. We give the full governing-equation derivations (Bosch–Hale reactivity averaging, adiabatic loss-cone confinement, the ambipolar Boltzmann relation, the Pastukhov end-loss law, the recirculation-limited gain, synchrotron reabsorption, the drift-cyclotron-loss-cone dispersion, central-cell interchange stability, and RF alpha-channeling), a GPU-accelerated computational methodology naming the real codes, and results across twenty-six figures and eight tables. A Saltelli–Sobol analysis confirms the plug ratio is the first-order driver of (first-order index S_1=0.41, total-order S_T=0.87); a 40 000-sample Monte-Carlo scan over the design ranges returns =1.317±0.14 with P(>1)=1.0, while a net-power uncertainty study near the operating cliff returns P(net>0)=73 %. We are explicit about a second binding lever: direct energy conversion is a declared, load-bearing input on a conservative-to-ideal ladder (ηdec = 0.49 conservative canon, 0.70 design basis, 0.85 proton-only ceiling), with a net-electric break-even near 0.52; the design point closes at the design basis and the conservative floor does not. The loss-cone microstability sits on the drift-cyclotron-loss-cone knee at a 6.25 % warm-plasma fill – identically 1/(n_p/n_c) – a maintained threshold rather than a passive margin, and a residual anisotropy-driven Alfv\'en-ion-cyclotron branch persists. The Aegis product wrap is stated quantitatively: a black-start whose direct-converter output crosses the start-up draw at ≈ 42 s, an islanded steady state at +104, a fleet capacity factor of 68–96 % across redundancy tiers, and a one-year grid-independent fuel package of order one tonne. Every net-power number is contingent on retiring the plug gate; the neutron-source performance and the low-activation siting case are not. No economic quantity is reported. Key results (frozen anchors): phid = 248.75 keV; Te = 89.7 keV; QE = 1.317 ±0.14; P_fus = 4298.5 MW; P_ch = 3868.0 MW; P_n = 233.95 MW. Methods & codes: CGYRO, OpenMC, GENE, Bosch–Hale. Live verification: 14 gate validator(s) with live-recompute cards (9 reproduced, 5 revised). See the Verification section and data/verification.csv. Related: Paper page · De-risking register · 3D model · Learn more about Kronos Part of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689). All numerical values are frozen design-point anchors; see the register. Public research artifact. No proprietary, financial, or supply-chain information is included.