FRC compression — pulsed field-reversed-configuration merging
Two counter-propagating compact toroids meet at the centre of a chamber, merge, and are compressed. SCPN-MIF-Core owns the device truth of the frc_compression_mif configuration and is the one member of the family with a release (v0.1.1, pre-alpha) and a portal of its own; this page places it in the family and points to it.
What the configuration is
A field-reversed configuration is a compact toroid with no toroidal field: a prolate ring of plasma current that reverses the external axial field inside its separatrix, so the plasma sits at unit beta in a field it largely generates itself. The FRC-compression scheme forms two such plasmoids at opposite ends of a chamber, accelerates them toward each other with a travelling magnetic field and merges them at the centre, where the merged plasmoid is compressed — by a coil, a liner or a jet shell — on inertial time-scales. The device problem SCPN-MIF-Core takes on is the kinematics of that meeting: two plasmoids must arrive at the centre phase-locked and spatially aligned within a window measured in millimetres and fractions of a radian, and the trigger that fires the compression must decide inside tens of nanoseconds.
What the core owns
- the pulsed-FRC kinematic surfaces: Doppler-corrected, distance-coupled Kuramoto synchronisation of the plasmoid pair (MIF-001), the chamber-fixed moving-frame remap (MIF-002) and the merge-window monitor (MIF-003);
- the pulsed-shot lifecycle as an adjacent transition ring — idle, ramp-up, flat-top, burn, expansion, dump, recharge, cool-down — with its guards (MIF-004), capacitor-bank dynamics (MIF-005) and Faraday recovery (MIF-009);
- the sensor path and the trigger: AER spike-buffer decoding (MIF-006), B-dot ADC to Q8.8 spike-rate quantisation with synthesisable SystemVerilog (MIF-007), the clocked, debounced single-shot trigger fabric and a registerless combinational fast-veto lane (MIF-008), kinematic safety (MIF-011) and plasmoid-merger Petri-net control (MIF-012);
- formal evidence on the open-source flow: SymbiYosys (Yosys + z3) k-induction proofs of veto dominance, single-shot and debounce safety, bit-true Python-versus-Verilator cosimulation (MIF-010, MIF-015), Lean proofs of the lifecycle and bookkeeping contracts;
- the binding to the SCPN Phase Orchestrator reactor registry and detection of the accepted SCPN-Fusion-Core FRC contract surfaces, without dispatching those physics kernels locally.
Declared targets and their status
| Target | Bound | Evidence today |
|---|---|---|
| Sensor-to-actuator latency | < 50 ns | cycle-budget relations proved formally (timing:cycle-budget-formal); post-route timing on a named device and measured full-path evidence remain hardware-gated — no artefact establishes the budget on silicon |
| Phase lock at chamber centre | |Δθ| < 0.01 rad | merge-window monitor, Python and Rust, upstream-pending |
| Spatial lock at chamber centre | |Δz| ≤ ±2 mm | same monitor; lock declared after three consecutive samples inside both windows |
| Plasmoid relative speed at merging | ≥ Mach 1 (v_z ≥ 300 km/s) | declared target of the kinematic model |
| Compression peak field | 20 T | declared target; no compression is solved in the core |
The carrier equations
The physics the merged plasmoid then obeys — non-adiabatic flux evolution after Ono et al. 1997 and magneto-Rayleigh–Taylor growth after Velikovich et al. 2007 — is implemented in SCPN-Fusion-Core and only detected, never dispatched, by the device core. The target plasma itself is described by SCPN-FRC-CORE: see the FRC section of the physics page and the explorer.