OPENCRETIN · THE OVERDRIVE LADDER · LIVE Λ-ITERATION — THE SOLVER RUNS IN YOUR BROWSER
Drive the lamp: candle → SUPERHEAVY
six device knobs. every change re-converges the trapped-field fixed point live (Zig→WASM kernels + long-characteristic transport, in a Web Worker). populations are the state; g589 is emergent; nothing is precomputed except the grey reference curve.
FUEL RATE — chemical power
STOICHIOMETRY φ — lean ↔ rich
PRIMARY DIAMETER
SECONDARY DIAMETER (0 = none)
NaCl PRESSURE AT THE WALL →
Na PRESSURE IN SECONDARY →
PHYSICAL KNOBS ONLY — everything else (T_wall, T_core, seed, blanket T/x, R, drive) is RESOLVED:
wall T = NaCl saturation inverted · blanket = Na saturation in the annulus gap · φ via heat-release factor [v1 ASSUMED — φ-resolved chemistry tables baking]
THE EMERGENT SPECTRUM — what the lamp emits, live · faced against the MEASUREMENT
left: ABSOLUTE emergent radiance, 500–1030 nm, synthesized from the converged populations through the cavity (I=(1−A)·Y/(1−A·X), diametral chord) — every line the kernel knows; the
self-reversed doublet emerges from transport (never drawn in). bars: out-of-band lines
(free escape). negative = net absorption at that wavelength (the trapped core).
THE MEASUREMENT — peak brightness, 6.7 suns, NaI flame, ND filter.
The model's job is to earn this curve. Scanner overlay = task #116.
THE SANKEY — one bar, every watt (live)
where the power actually goes — real flows vs the photon tank (live ledger · DEVICE TOTALS for the H = 1 m column · 2-D slice model)
pump IN
—
what the power system pays — device total (lower bound)
→ heat (quench)
—
collisional de-excitation → gas → thermal loop
→ 819 nm + 330 nm photons OUT
—
out-of-band lines: pass the dichroic — THE HARVEST ·
→ 589 nm net OUT
—
in-band leakage past the 0.97 mirror
⟳ 589 TANK
—
gross recirculating power — a STOCK, not a flow; costs nothing steady-state
ledger closure
—
in − heat − rad, % of pump (pooling unbooked)
the trap is a photon tank circuit: the recirculating 589 power is the reactive/flywheel
component (can exceed the pump by orders); the drive pays only the dissipative sinks —
quench heat (recovered thermally) + photons that actually leave (819/330 through the dichroic,
3% in-band leakage). 589 is the reservoir; 819 is the tap.
THE EMERGENT MIRROR — A(λ) of the sodium blanket · R is an OUTPUT, not a knob
the secondary holds sodium metal: its vapor (saturation at T_metal) forms an
x-thick resonance-scattering blanket — optically thick exactly AT the D lines, transparent out of band.
τ₂(λ)=κ₂(λ)·x from the same lines kernel as the primary; two-stream slab albedo; buffer pressure sets
quench loss (ω = A₁₀/(A₁₀+k_Q₂)) and broadening sets the band WIDTH. Sodium mirrors sodium light.
Crank T_metal / x / buffer and watch the band grow — this and r_p are the 50-OOM geometry levers.
THE ANNULUS — device section, live: T(r), photon occupation n̄(r), and what escapes at the mirror
left: the real cross-section — gas column colored by local T, sapphire wall (n=1.77) at 40 mm,
vacuum annulus 40→46 mm, band mirror (R=0.97, 555–660 nm) at 46 mm.
right: T(r) against the left axis, n̄(r) — 589-mode photon occupation —
against the log right axis, drawn over the same geometry. At the mirror: the live escape flows.
Reflection off the coaxial mirror preserves the impact parameter → each chord is a closed cavity:
I_in = R·Y/(1−R·X). [sapphire thickness drawn 1.5 mm, schematic]
the ladder — wall & core b vs drive · grey = reference (1 atm, R=0.97, seed .008) · dot = YOUR knobs, live
radial profile — b(r) across the column, live solve
overdriven band: b ≈ D / (k_Q·f₃ₚ) — pump vs quench (k_Q ∝ P: pressure knob moves the whole band) ·
ceiling: n₃ₚ/n₃ₛ → g₁/g₀ = 3 (the drive's chemical potential, NOT kinetic T) ·
trap R feeds n̄ (field re-pump) · seed sets the column depth (trapping lid) · wall T sets the saturation cap + skin vapor ·
top rung: κ < 0 — population inversion, D-line gain inside the trap cavity: the lamp lases.
