NORTH POLE EXPRESS 24 — INDEPENDENT DIGITAL QA, REVISION B
Date: 2026-11-02
Reviewer: /root/qa_review, independent of the geometry authors.
Scope: geometry container, removable-fin mechanism, coupon/display accessories,
assembly reference and local diagnostic slicing. No geometry was repaired by QA.

DISPOSITION
PASS for the bounded digital geometry, nominal-interface, service-motion and
package checks below. The file contains all 14 printed physical parts. One
coupler end-feature finding, QA-B-01, was corrected by the designer and rechecked.
Physical PETG fit, printed strength, sealing, heat exposure, recovery and flight
suitability remain untested. This is a prototype geometry delivery, not an
approved print job, validated AMS profile, universal fin standard, or flight release.

FINAL REVIEWED IDENTITIES
Main: exports/B/North_Pole_Express_24_ALL_PARTS_B.3mf
SHA256: 977a1e07b70d69385f498c9def8cc6cf4cae7e5f9f77473ab693d71de0f20d6e
Assembled: exports/B/REFERENCE_ONLY/North_Pole_Express_24_ASSEMBLED_B.3mf
SHA256: 7416601630a029cea16d38d4a51e1dbd5c7b59c3c092129729298fe5a5cb8dd3
Specification SHA256: 0b27701ea7343cc5bb0b729e1ea7236107ee301d504d70777413bd0e9a2368f1
Main manifest SHA256: c07188a844d91a26c2fa72e70587e8220850330d21a0f39b662c2fa05266f0cd
Exact source and individual-part hashes: QA_FINAL_IDENTITY_B.json.
QA_PACKAGE_B.json holds final dimensions, volumes, topology and placement data.

Tools: Python 3.11.9; numpy 2.4.6; trimesh 5.2.0; manifold3d 3.5.4;
CadQuery 2.7.0; shapely 2.2.0; Bambu Studio 02.08.02.61.
Scripts: qa/check_revision_b.py, qa/check_supplement_b.py, and shared read-only
helpers in qa/check_independent.py. Units are millimeters unless stated otherwise.

1. Completeness, topology and package — PASS
   NP-01/02 body sections; NP-03 keyed coupler; NP-04 nose/shoulder;
   NP-05 nose anchor; NP-06..09 four removable fins; NP-10/11 motor rings;
   NP-12/13 rod guides; NP-14 aft keeper. All 14 appear once on the build plate.
   Purchased motor tube, hook/thrust ring, screws, heat-set inserts, recovery gear,
   adhesives and motor are external items, not missing printed parts.

   Every STL is one connected, watertight, consistently wound positive-volume
   solid with finite coordinates, zero degenerate triangles and two faces per
   edge. Final triangle count is 1,835,030. Manifest bounds/volumes/hashes agree.
   Core 3MF resources, transforms, millimeter units and readback were checked.
   Print-object bounding boxes do not overlap. Final placement is approximately
   X8..216.5326, Y8..170.4156, Z0..240, within the declared 230x250x250 envelope.
   Surface color references are valid. Color data does not establish filament
   assignments, AMS slots, purge space or the owner's actual process settings.

2. Dimensional intent and assembled reference — PASS
   Decorative airframe length is 610; the keeper extends to Z-2.4, making the
   printed assembly 612.4 tall before motor and screw-head projections. Fin span
   remains 145. Base body OD is 45; the peppermint relief extends beyond it.
   Body sections are each 240 long; the visible nose is 130 plus its 25 shoulder.
   The fin face begins at radius22.55, adjacent to the radius22.5 nominal skin.
   The captured T-foot is inside the body, at radius18.6..20.4.

   The assembled 3MF retains all 14 original mesh and color XML payloads
   verbatim. They are identity-transformed components of assembly object100,
   with one build item translated128,128,2.4. Bounds are55.5..200.5 in X/Y and
   0..612.4 in Z. This is a multipart reference, not a one-piece printable build.
   Evidence: ASSEMBLED_3MF_CHECK_B.json.

3. Nominal interfaces and continuous service motion — PASS
   Nominal body/fin, keeper/fin, ring/body, coupler/body, nose/body, anchor/nose,
   guide/body and assumed26 OD paper-tube interfaces have no material overlap.
   Numerical boolean tolerance is0.05 mm3; detected screw-envelope residuals
   were below0.000001 mm3. Corrected coupler/body pairs are exactly zero.

   Each complete fin was conservatively projected onto XY and extruded through
   its entire swept Z interval from140 aft of its seated position to seated.
   This envelope contains every intermediate translated position, not only a
   set of samples. All four envelopes clear the fixed body, both motor rings,
   guides, assumed paper tube and diagnostic hook envelope.
   The keeper's two height profiles clear a30 mm straight-aft withdrawal.
   Both nominal screw/head envelopes clear seated and withdrawing positions.
   Positive-contact probes confirm geometric radial capture, forward stops and
   keeper retention; these probes do not determine load capacity or stiffness.
   A continuous4.8 diameter rod envelope at radius28.5/225 degrees is unobstructed.

   Screw stack assumed: M3x6, no washer, head<=8 diameter and<=4 high;2.4 keeper
   thickness leaves3.6 projection into a5 deep blind pilot. Nominal tip clearance
   is1.4. Actual inserts/threads, thumb access, engagement and pullout remain
   physical checks. The hook envelope is a diagnostic assumption, not a verified
   representation of every purchased commercial hook.
   Evidence: QA_MECHANISMS_B.json and QA_COUPLER_ENDS_B.json.

4. Permanent cavity walls and minimum features — PASS for tested geometry
   With fins and keeper absent, interior test volumes show continuous1.2 nominal
   track backs, side walls and closed forward caps, and solid blind-boss roofs.
   Track pockets/mouths are empty through the intended stroke. Ring notches do
   not cut those fixed walls. This establishes modeled material continuity,
   not airtight printed walls or sealed adhesive joints. Permanent ring/body/
   cassette bonds and the paper motor tube remain part of the gas boundary.

   QA-B-01: the initial coupler backing taper produced a0.28-thick tip and a
   separate0.086-thick strip at Z220.15, with a radial gap between them; the
   upper end was symmetric. This contradicted the stated1.2 minimum backing.
   The designer replaced the taper with a continuous full-length backing box.
   Final sectional and volumetric checks establish a continuous1.2 center web
   through both ends and the span, while clearing both body sections. CLOSED
   for digital geometry. Check recovery packing past the local inward rib.
   Prior failure evidence: QA_COUPLER_ENDS_B_PRE_CORRECTION.json.

5. Fit tests — PASS for digital contents and nominal motion
   FIT_TESTS/01_SHORT_CLEARANCE_B.3mf: four parts, female clearances0.15/0.25/0.35
   per mating face plus one common male. All three include the closed front cap.
   FIT_TESTS/02_FULL_LENGTH_TRACK_FIN_B.3mf: two parts, actual NP-06 fin and
   full-length default0.25-clearance track. The fin matches NP-06 exactly and
   the track is a subset of the actual lowerbody track.
   FIT_TESTS/03_TAIL_KEEPER_FUNCTION_B.3mf: six parts,35-long tail, actual NP-14
   keeper and four short test levers. The keeper matches NP-14 exactly.
   All12 coupon solids pass topology checks, all three layouts are separate and
   within the declared envelope, and the tested insertion/keeper interfaces clear.

   The long set has no lock. The short-tail set checks keeper operation and
   short-root rocking. Neither alone validates a locked full-length joint.
   Select a clearance from actual prints, verify full-length sliding, then
   assess the actual full fin locked in the full body, including repeat use and
   tip movement. Changing clearance requires consistent regeneration, not scaling.
   Evidence: QA_COUPONS_B.json; procedure: FIT_TESTS/PRINT_FIRST_README.txt.

6. Optional display fillers — PASS for bounded digital checks
   Four connected closed solids, valid color references, no body/keeper overlap,
   clear conservative insertion envelopes and nonoverlapping plate placement.
   The four fillers replace fins only for display and are not flight components.
   Archive SHA256: c3298b8002d78c4e626b57ef7fc3752656356be4285d8f60c101a6c919de09fe
   Evidence: QA_DISPLAY_FILLERS_B.json. Physical fit and finish remain untested.

7. Diagnostic slice — PASS for bounded process evidence
   The designer ran a local H2S0.6 / stock PETG Basic monochrome proxy with
   Arachne,0.18 layers,0.3 first layer, no supports and no brim. QA reads the
   emitted G-code only; no printer transmission was performed. The final source
   hash matches the main file above. Slicing returned0 with no warning message.
   Independent G-code reading confirms extrusion for all14 objects and body
   paths through the cassette-cap height interval. Diagnostic consumption is
   196.304 g and estimated time is about6.49 hours; these are process estimates.
   Final evidence: qa/slice_B/summary.json and QA_SLICE_COVERAGE_B.json. Prior
   slice evidence is retained in qa/slice_B_before_coupler_fix/ and the prior
   coverage JSON, rather than being presented as the corrected geometry.
   Emitted paths do not validate bead adhesion, thin-wall strength, support
   adequacy, sealing or the actual multicolor process. The G-code remains QA
   evidence and must not be treated as a prepared printer job.

8. Change control and remaining physical work
   The final correction changed only NP-03. Other13 STL hashes, original 3MF
   mesh payloads, plate transforms and color palette remain identical to the
   first B baseline; prior mechanism/coupon/display results therefore remain
   applicable. Coupler topology, web, body fit and final package identity were
   rechecked. A evidence remains preserved and is not reused as B flight approval.

   Solid-volume plastic estimate is210.793 g at assumed1.25 g/cm3, before all
   purchased hardware, recovery items, adhesives and any secured ballast. This
   is not loaded mass or a measured print. E12 suitability/delay, final balance,
   actual motor-mount dimensions, support/brim strategy, PETG coupons, locked
   full-length play, repeated wear, insert installation/retention, sealing,
   structural loading, thermal exposure and recovery remain unvalidated.
   Nominal0.5 axial clearance is intentional prototype play; no zero-play,
   clamped-joint or quantified reuse-life claim is made.
