Stability-Problem Map Across Both Machines

The same variational stability question specializes to different operators on the breeder and the burner; this map pairs each mode with its equation and its AI treatment.

One principle, machine-specific operators

Every stability limit in the stack descends from the ideal-MHD energy principle: an equilibrium is stable when no allowed displacement lowers the potential energy. What changes between machines is the geometry and therefore the operator, the dominant modes, and the honesty caveats attached. This page is the cross-reference the twin uses to route a stability query to the right module.

Breeder (Hyperion) modes

• Vertical n=0: rigid-shift eigenproblem; active PF feedback; wall-time growth.
• External kink / RWM n=1: delta-W with wall; slow feedback-controllable growth.
• Ballooning high-n: 1D s-alpha ODE; bounds edge pressure gradient at delta -0.30.
• Tearing / NTM: resistive, rational-surface islands; q-profile dependent.

Burner (Aegis / MetroVolt) modes

• Interchange / flute: average curvature; needs minimum-B magnetic well.
• Mirror-mode / anisotropy: driven by p_perp > p_par pressure anisotropy.
• Loss-cone / microinstabilities: velocity-space driven, kinetic not fluid.

Which math, which caveat

The table below is the decision the twin encodes: given a machine and a mode, which mathematical object is solved and which honesty gate travels with the answer. The burner rows always carry an extrapolation or feasibility caveat because no comparable device exists.

Nothing in this map claims hardware net gain; all breeder limits are simulation-validated pending FOAK ~2030, and all burner limits carry the four gates. The map exists so no query is answered by the wrong operator or without its caveat.