HexCell — the LightCell tower unit, exploded
one printable alumina hex tile, 150.7 mm flat-to-flat · 101.6 × 1257 mm quartz tube · 315 cells per 420 m column · drag the assemble slider, hover assembly steps, every dimension traces to HEX-CELL.md + COMPONENTS.md
Provenance banner. Component design per HEX-CELL.md + COMPONENTS.md. [T] textbook value, confirm against spec sheet · [UNSOURCED] estimate needing derivation · [OP] operator datum. No fabricated thresholds — condensation is transport, not a limit.
Component list (one cell)
C1 · Hex header, bottomalumina SLA · seat+salt+injector · 150.7 mm OP
C2 · Hex header, topalumina SLA · sliding seat · 150.7 mm OP
C3 · Outer quartz tubefused quartz · 101.6×1257 mm · wall ~2.5mm · $49.62 OP
C4 · Inner PCA emitterpolycrys. alumina · sprung mount · ⌀≤100mm U
C5 · Optional sleevevac jacket / ITO hot-mirror U
C6 · Honeycomb columnextruded alumina · top→bottom T
C7 · Salt chargeNa reservoir salt, comp. TBD U
C8 · Springs (cool end)Inconel · ~10–20 N preload U
C9 · Gaskets (cool end)compliant, cool ends only
Cost stack: quartz $15–25 vol · alumina print <10% BOM (OP 4× below retail) · honeycomb cheapest ceramic extant · order $100–200/cell before internals + silicon.
Assembly sequence (COMPONENTS.md §4)
  1. C1 bottom header: salt charge into seat bed; injector flow-test cold.
  2. Drop C3 quartz into cradle; verify float — no radial bind.
  3. Insert C4 inner emitter on sprung mount; optional C5 sleeve.
  4. C2 top header on; C8 springs preload from cool end; C9 gaskets cool ends only.
  5. Cold leak/flow test; log per-cell serial. Cassette-level burn-in before hoist.
Tiles clip edge-to-edge. The hex tessellation is the structure — no cassette frame (that concept SUPERSEDED 2026-07-19). Bad tile → unclip, lift out vertically, refurb on bench.
LIVE PHYSICS · v3a cell-scale — loading…
cell solver (v3a): fixed-point Pdep, no clamp, single-bounce greybody enclosure. True cell geometry — L=1257 mm, R up to 50 mm bore. Kernel heatmap uses bench-scale na_grid2d (R≤40 mm, L≤400 mm — its validity cap).






—
T(r,z) — na_grid2d.zig bench-scale kernel (R≤40 mm, L≤400 mm). Cell solver above runs true geometry. —
ledger: P_escape / P_wall / P_quartz — total ≡ Pin at convergence
conservation residual: —
emergent spectrum — cell_solver.js (v3a) · fixed-point Pdep (no clamp) · greybody escape gated · [MODEL v3a single-zone]
Assemble drag: exploded → nested
0.00
Hover a numbered step (left) to highlight parts. Callouts sit at exploded positions; at slider=1 the stack collapses to seated geometry. Lengths and diameters are true-ratio (tube 12.4:1 L/D T), tile pitch 1333 mm OP.
Cross-section · looking down the bore
Hex flat-to-flat 150.7 mm · side a=87 mm · inscribed ⌀ 150 mm · quartz OD 101.6 mm · annular margin ~24 mm · footprint 0.0197 m². Emitter side area / cell = π·0.1016·1.257 ≈ 0.40 m² = 20× vertical gain vs footprint.
alumina hex (C1/C2)
quartz tube (C3)
honeycomb (C6)
PCA inner (C4)
Thermal — the governing mechanical problem
CTE mismatch — quartz 0.55 vs alumina 8 ppm/K T. At ΔT ≈ 800 K over 1257 mm: quartz grows ~0.55 mm, alumina frame ~8 mm. → fixed seat at bottom, sliding seat + spring at cool top. Radial across 150 mm: alumina ~0.9 mm, quartz ~0.07 mm — never clamp quartz radially hot; salt bed = compliance.

Links

../dyson-cylinder/ · live sim
HEX-CELL.md · unit doctrine
COMPONENTS.md · parts + assembly
../basis/ · na_grid2d recovered sim
Cassette → column → tower · cells hang, never load-bearing
interlocked tile plates (tiles clip edge-to-edge) → 315 tile-rows per 420 m column OP · 42 m ⌀ tower cross-section 1385 m² · annular shell fill ~27% OP, not solid pack · per-plate burn-in on ground before hoist. Cell mass ~5–8 kg U · ~2 t/column · tower structure supports plates at intervals; tessellation is the diaphragm.
HexCell explorable · dark #070809 / IBM Plex Mono · every number sourced or flagged · 2026-07