# TASK B - vertical circulation audit (INT R02 vs SHELL C02), read-only

Sources: layout.json (INTERIOR_LAYOUT), repo 4ce28d8 (ROOM/DOOR_SCHEDULE, MODELING_BRIEF, COMPONENT_SCHEDULE, OPEN_DECISIONS, QUIK_WORK_ORDER, EXT R06 DesignTargets), C02 GLB sliced with refit.py helper code (copied via exec, unedited).
Labels: measured / proposed / interpreted / not_testable. Reproduce: `C:\q2\v\Scripts\python.exe C:\q2\out\B_vertical\audit_b.py`.

## What exists (proposed)
Same 5 core rooms on each of L00, L01, L02, L03, L25, L26 (ids LXX-RM090..094):
- RM090 Stair A  x -18..-10, y -8..6 (8 x 14)
- RM091 Stair B  x 10..18, y -8..6 (8 x 14)
- RM092 "Three passenger lift shafts" x -6..6, y -8..-3.2 (12 x 4.8); 3 lift door records D086/D087/D088 at x=-4/0/4, y=-8, 1.5 x 2.4 m
- RM093 "Service lift / MEP reserve" x -10..10, y -3.2..2 (20 x 5.2); notes/MD05 say shaft is x 6..10 (4 x 5.2); door D090 (8,2) 1.8 x 2.6 to RM094
- RM094 Service vestibule x -10..10, y 2..6; D091 (8,6) 1.5 x 2.7 to cross corridor RM083
- Stair doors D084 (-18,0) and D085 (18,0), hinged 1.5 x 2.7, axis y, on the outer long side of each stair at mid-length.
Counts per level (all 6 levels): 2 stairs, 1 lift-shaft room with 3 lift door records, 1 service reserve. Matches "3 lifts + 2 stairs + service reserve" only by name/door count; no individual lift shaft rooms exist.
Proposed component targets: cab 2.8 x 3.2 x 2.7 (MD04); cell 4 x 4.8; service shaft 4 x 5.2, door 1.8 x 2.6 (MD05); stair 1.8 m clear, 0.30 tread, two flights, 36 risers/6 m, 28 risers/4.5 m (MD06, brief).

## Continuity (measured on proposed data)
Footprints are identical on all six levels: dx=dy=dw=dd=0, IoU=1 for L00-L01, L01-L02, L02-L03, L25-L26 (core_continuity.csv). No core room overlaps any other room on its level (edge contact only). L03 and L25 footprints are also identical, which makes a straight stack CONSISTENT, but:

**Continuity through floors 4-24 (L04..L24, z=24..120) is NOT evidenced by the layout.** 39 of 45 floors have no rooms, doors or cameras. The brief's "interpolate through intervening floors" is an instruction, not evidence. Not testable.

## Rise and clear height (measured / arithmetic)
Rises (EXT R06 slab list): L00->L03 3 x 6.0 = 18; L03->L09 6 x 6.0 (to z=54); L09->L25 16 x 4.5 = 72; L25->L26 4.5. L03->L25 = 108 m, 21 unlaid-out levels. Core clear_height_m 5.4 (L00-L03) vs rise 6.0 (0.6 build-up); 3.6 (L25/L26) vs 4.5 (0.9). The core room clear height is below one rise, so stairs/shafts need the upper slab open (README: OBJ omits stair/shaft floor plates; stairs not modelled).
Riser from proposed counts: 6.0/36 = 0.1667 m; 4.5/28 = 0.1607 m. Risers needed at proposed counts: L00->L03 108; L03->L25 6x36+16x28 = 664; L25->L26 28.

## ASSUMPTION SWEEPS (not design facts) - stair_assumption_sweep.csv, stair_headroom_sweep.csv, stair_door_vs_assumed_flight.csv
- Riser 0.15-0.19 (assumed), treads = risers-1 per flight (assumed), landing depth 1.2/1.5/1.8 (assumed), tread 0.30 (proposed). Plan length of a two-flight run of 18 risers: 7.5-8.7 m; fits the 14 m Y extent, not the 8 m X extent at 1.5-1.8 landings. Width 2 x 1.8 = 3.6 of 8.0 m.
- Headroom (assumed 1.9-2.2): 6 m rise mid-landing headroom 2.40 m (ok); 4.5 m rise mid-landing headroom 1.35 m (below every assumed value) unless the L25/L26 upper slab is open above the landing and all of flight 2 (OBJ omits plates, not modelled).
- Door vs landing: D084/D085 spans y -0.75..0.75, i.e. 8.0 m from the -y end, 6.0 m from the +y end, 1.0 m off centre-line. If the flight is anchored at the +y end the door span overlaps a flight (e.g. y -0.9..4.2 for 18 risers at 1.8 landing): 10 of 24 swept combos overlap. Risk only; real stair placement unknown.

## Doors / conflicts (measured)
No door off its shared edge, none crossing a core interior, none touching a core room it is not connected to; assumed full-disc swing (radius = clear width) of hinged doors intrudes into no unconnected core room. Stairs are not entered from the lift hall (no door), only via RM081/RM082. Lift door 1.5 x 2.4 is 0.30 m lower than the 2.7 m cab target. RM093 (20 m wide) is back-to-back with RM092 over 12 m with no door; the service route is RM093 -> D090 -> RM094 -> D091 -> RM083.

## Shell and rib fit (measured, C02 mesh; core_shell_rib_fit.csv)
Min wall margin / rib clearance over core rooms 090-094: L00-L02 28.52 / 34.45 m (Podium_wall_1; ribs R5), L03 19.00 / 31.52 (Podium_wall_2; R3), L25 14.10 / 18.00 (skin; R3 at z=129.6), L26 13.94 (L26-RM090) / 17.42 (L26-RM091 vs Rib_R4). No intersection anywhere. My margins agree with published C02_FIT_COVERAGE (see diff column). A straight extrusion of the footprints through slabs z=18..126 stays inside the C02 wall/skin: min 12.0 m (Podium_wall_3, z=36, interpreted). C02 contains no core, shaft, stair or slab meshes, so this is plan-vs-skin only.

## Capsule sweep 0.6-1.0 m (capsule_sweep.csv; capsule UNKNOWN)
Door records: 1.5 m (18 passenger records), 1.8 m (6 service records), 1.2 m service door (MD01), stair 1.8 m clear. Width clearance per side 0.25-0.45 m (1.5 m door), 0.40-0.60 m (1.8 m). All pass by width; sill, height, step-up and door timing untestable.

## Not testable
Lift pit/overrun/machine room/speed, cab door swing, stair landing depth/handrail/stringer geometry, service cab size, continuity L04-L24, structure/MEP intrusions, headroom at real stair geometry, movement capsule/camera envelope (unspecified), engine movement.
