# Spiggen II model evidence — 2026-10-14

Scope: late museum configuration; a parametric reconstruction for a keyboard-and-mouse museum kiosk prototype. Sources are the supplied research packet and actual cube images, inspected locally. These are not factory drawings or a photogrammetric survey.

## 1. Reference priorities and fixed anchors

- A01 `derived/extracted/signallinan34-p20-img1.jpeg` (1194x827), S02/S03: clean Granholm longitudinal cutaway + horizontal plan. Nose RIGHT in both. Best source for longitudinal placement and pressure-compartment relationships.
- A02 `sources/museum-technical-poster-3922616.webp`, S05: same layout explained with tank/system labels. Not independent evidence for every line of A01. Its `Bredd: 11,94` label is an error on a LENGTH arrow (C44).
- A15 `sources/commons-spiggen-tore.jpg`, S12: detailed late exterior side view, nose RIGHT; use perspective camera matching, not pixel-width-to-metre scaling. Original 5344x3008. CC BY-SA 4.0 option recorded in packet.
- A04 `sources/museum-submarine.jpg`, S01: late museum bow quarter, valuable for round bow cross-section, dorsal panels, plane mount and sail front.
- A05 `sources/propulsor-3922566.jpg`, S04: three swept propeller blades inside a separate thick annular shroud/nozzle. Thin upper members are structure, not extra blades. Nozzle rotates in two planes (C19/C20); propeller rotation must remain independent.
- A06 `sources/hydroplane-3922852.jpg`, S04: broad, short, rounded-edge rectangular bow plane high on the side with a visible mount/support. Added in later life; absent on the earlier drawing is not permission to omit it from the museum configuration.
- A07 `sources/museum-controls-3922865.jpg`, S04, plus A08 `sources/cockpit-3922855.jpg`: actual late controls and two adjacent pilots.
- A19/A20 `sources/norman-slide-84.jpg` / `85.jpg`, S09: low-resolution stern/bow confirmation only; do not infer precise dimensions.

Published anchors (medium-high source confidence, not a current survey): total length 11.94 m (C08/S06), pressure-hull diameter 1.70 m (C09/S06), total height 2.60 m (C10/S06; mast/height datum not resolved). Exterior beam and plane span remain unknown (C11). Do not set exterior beam to 1.70 m as a fact.

Runtime model axes: X forward, Y up, Z starboard. For the table below, t=0 is the aftmost nozzle rim and t=1 the bow tip; x=(t-0.5)*11.94 m is a convenient centred modelling coordinate. This origin is a software convention.

## 2. Approximate longitudinal stations

These are visual traces from A01, rounded deliberately. They are editable reconstruction priors, NOT measured dimensions. Typical placement tolerance at this stage is at least +/-0.02 of length; no claim that a feature was surveyed to centimetres. A01 whole drawing runs approximately px92 to1082. Its separate 5 m scale bar is about417px, giving ~83px/m, reasonably agreeing with 11.94 m over the drawn overall length. Vertical/horizontal image distortions and drawing conventions remain.

| Region/node | Approx. t | Visual basis / interpretation |
| --- | --- | --- |
| Nozzle annulus | 0.00–0.05 | Thick annulus separate from aft fairing, A01/A05 |
| Bare shaft / articulated support | 0.04–0.10 | Do not connect propeller and nozzle as one rigid spinning assembly |
| Outer stern taper | 0.07–0.25 | Long conical/rounded transition to full hull, A01/A15 |
| Aft pressure end / aft trim-regulating tank | 0.20–0.27 | Rounded pressure end and blue tank volume in A02 |
| Diesel and machinery region | 0.27–0.41 | Main diesel low along shaft line; surrounding service machinery |
| Main battery block | 0.41–0.51 | Low dark rectangular mass immediately aft of central lockout; electrical box above |
| Central lockout chamber | 0.51–0.64 | Bulging middle pressure volume beneath entry trunk; opposed round internal hatches |
| Entry trunk/hatch | 0.54–0.60 | Vertical cylinder to sail top, not a glass cockpit |
| Two pilot seats / working area | 0.64–0.74 | Side-by-side across beam, both face bow, not tandem |
| Forward pressure end / trim-regulating tank | 0.76–0.83 | Rounded end and blue tank volume in A02 |
| Forward pressure vessels / emergency battery region | 0.84–0.92 | Several upright bottles in the outer bow space; distinguish functions |
| Rounded bow closure | 0.87–1.00 | Blunt rounded nose, not a pointed torpedo cone |
| Sail envelope in A01 | 0.40–0.61 | Long aft ramp, rounded top then much steeper short forward drop; tune to A15 perspective |
| Lifting eyes | approx.0.35 and0.70 | Small triangular arch eyes on dorsal structure; not generic repeated rings |
| Late bow planes | approx.0.82–0.86 | A15 photo prior only; measurement required for axial position, chord and span |

Shape priors: A01 external midbody depth is roughly 1.9–2.0 m if its vertical drawing scale is accepted; nozzle outer diameter roughly 1.3–1.5 m. Treat BOTH as inferred adjustable values, not published characteristics. Pressure cylinder radius 0.85 m is the only published radial anchor. Midbody should be elongated and fairly constant in section, with a raised dorsal fairing/sheer strip; it is not just a capsule with a tower pasted on. Exterior cross-sections need later survey.

Sail must read as low and integrated with dorsal fairing. Its aft side ramps gradually up from deck, forward side falls steeply to deck, with rounded corners/edges. A15 shows an open top hatch near the forward part and Swedish three-crowns roundel on the side. White rectangles on A04 hull are sunlight, not camouflage markings. Do not bake them into material. Surface is matte dark charcoal/black, worn panels/rivets/seams; not shiny blue steel. The museum blue cradle is omitted at the user's request. The long lower frame is also omitted following the user's later request; this display choice is distinct from the source evidence.

Hatch correction: A28 (1. Ubåtsflottiljen, via Ny Teknik) shows a thick circular lid nearly upright aft of the opening. The late tower/u panorama shows its inner wheel and linkage. Aft hinge position is inferred from the open pose; hinge dimensions and exact angle remain unmeasured. The renderer uses one shared lid, open in museum views and closed throughout simulation. See `interior-comparison.md` for links and the distinction between older service and late museum views.

## 3. Interior reconstruction from real panorama tiles

Dataset `panorama-export/tiles/uboat-*`: one 512px cube face per direction at `1/{b,d,f,l,r,u}/0/0.jpg`; each preview is a six-face strip, not equirectangular imagery. Original README correctly says five folders but only FOUR distinct captures exist.

CRITICAL: `uboat-aft` is byte-identical to `uboat-airlock`: preview AND all six cube faces have matching SHA256s. Do not display it as a fifth independent aft station. Alias or exclude it. Example preview SHA256 both: `9cbf0cf33d45278af4f2f7fc271bd132d4514a9a3398256b28d850b0a2fbaf17`.

Actual inspected face directions:

| Scene | f | r | b | l | u / d |
| --- | --- | --- | --- | --- | --- |
| uboat-control | port fuse bank/depth controls | forward central monitor stack | starboard electrical panel and seat | aft round hatch into airlock | ceiling piping / two seats and low floor |
| uboat-airlock | aft hatch to machinery | ladder and lamp wall | forward hatch with visible cockpit | both stowed round covers | vertical entry trunk / oval grating between strapped cylinders |
| uboat-engine | forward round hatch to airlock | side equipment | aft fan/equipment end | side equipment | ceiling / machinery-cover surface |
| uboat-tower | cylindrical trunk wall | cylindrical trunk wall | cylindrical trunk wall | cylindrical trunk wall | open top showing museum roof / airlock and ladder below |

All 24 faces of the four unique stations were inspected individually for this refinement. The per-face comparison and resulting equipment corrections are recorded in `interior-comparison.md`. Global yaw and precise dimensions remain visual estimates; tower global yaw has no surveyed calibration.

Navigation directions from original `92adb27:lib/uboat-tour.ts` support the visual identification: airlock -> control yaw3.1149, airlock -> engine yaw0.0508, airlock -> tower pitch-1.3565; control -> airlock yaw-1.86095. Use original link geometry only as a starting point, not scene physical station coordinates.

The panorama viewer preserves each capture's local cube orientation. Its face mapping is verified visually for readable text and continuity. Equipment is placed separately in the vessel frame using observed hatch relationships; cubemap yaw is not treated as surveyed physical coordinates.

Visual detail to model: off-white curved pressure shell; dark almost black/green equipment panels; dense silver/grey overhead pipes with black round valve wheels and red lever handles; two dark cushioned seats side-by-side; central forward stack of rectangular monitors (dark in museum), switches/buttons below; angled fuse banks left; large dial cluster on left and some electrical meters on right. Seats face a closed instrument wall. There is NO forward optical window.

A07 controls: depth wheel at upper-left of compass; course wheel below compass with a spinner knob; compass is large white circular card; course rudder-position indicator above it; depth-plane indicator upper right; propeller control on right labelled PROPELLERMOTOR with FRAM/BACK; horizontal upper row contains small displays/autopilot. Gauge face +/-40 marks do not establish real motion limits. Retain these labels/relationships where visible but do not reconstruct unknown circuits.

Central chamber: white round end walls, opposed circular hatches with thick rims and open covers; covers can have a small viewing port, which is an INTERNAL hatch feature, not a forward exterior window. A vertical off-white trunk and ladder lead to the sail hatch. No 'periscope' should be added from generic submarine expectations. Engine space is extremely crowded: service boxes high on both sides, pipes/cables along crown, large low stainless/grey machine covers, visible fan at aft end. The panorama does not show a clean isolated full diesel; diagram gives its functional location.

## 4. Source text exclusion / fidelity rules

Do not copy narrative/Q&A from original `lib/uboat-tour.ts` into the kiosk. Actual text at source commit92adb27 claims 24 crew, three weeks endurance, fifteen-millimetre hull, periscope, and involvement in 1983. These conflict with the supplied research packet or are unconfirmed. Use only imagery and geometrically validated hotspots from that legacy export. Research packet C25 supports two pilots plus1–2 divers; C32 supports paired forward pilot positions; no AIP/Stirling, torpedo tubes or forward window should be invented.

Keep exterior form hull distinct from pressure hull. Three pressure compartments do not mean three identical externally separable modules. Keep fore/aft main ballast separate from trim-regulating tanks, fuel, bilge and pressure vessels. Rendering pressure-shell thickness is a visual setting until documentary evidence arrives. Any water filling animation must apply to named tanks, not ordinary living volume.

Every substantial object should carry source_ids, asset_ids, confidence, measured_or_inferred, version. Suggested version='late-museum'; source='A01 layout, A04/A15 exterior, panorama interior' with dimensions classified as published or inferred per above. Unknowns remain replaceable configuration parameters.

## 5. Acceptance views

1. Orthographic starboard/port longitudinal profile against A01; length11.94, low sail, blunt bow, long aft taper.
2. Top cutaway against A01/A02: paired pilot seats forward of central chamber, machinery/battery aft; no mirrored equipment assumed without reference.
3. Museum camera match against A15 and A04 for silhouette and plane mount.
4. Stern view against A05/A19: exactly3 blades, separate thick annular nozzle and articulated support.
5. Cockpit interior against actual control cube: correct two-seat scale and enclosed instrument panel.
6. A labelled reconstruction layer for unknown dimensions, hidden piping, beam and simulation coefficients; no claim of completed exact replica before survey/museum review.

All A01/A02/A04/A05/A06/A07/A08 rights remain unconfirmed in the packet. They can inform local modelling; public display inside the distributed kiosk needs the packet's rights check. The existing project panorama's capture ownership is not documented by README; avoid inventing an attribution.

## Управление глубиной: повторная проверка 2026-10-14

Проверена сохранённая полная статья S04: Angelica Söderberg, Teknikhistoria, 30 июня 2023, интервью с конструктором Mats Nordin и служившим на лодке Roger Lundkvist.
Источник: https://teknikhistoria.nyteknik.se/farkoster/har-ar-historien-om-svenska-flottans-hemliga-miniubat/3922552.

Nordin сообщает, что предусмотренные проектом рули глубины исключили из раннего прототипа ради бюджета. Lundkvist описывает хорошую управляемость по курсу, худшую по глубине и своё участие в разработке последующих рулей. Статья и фотография A05 подтверждают отклонение карданной насадки вертикально и горизонтально; A06 показывает поздний носовой руль.

Следствие для музейной версии: нужно показывать и носовые рули глубины, и отклоняемую насадку. Вращение винта само по себе не является командой глубины. Источник не устанавливает точную связь приводов, углы, распределение усилий и работу автоматики позднего исполнения.

Исправления прототипа: оба носовых руля теперь вращаются вокруг собственных неподвижных осей. Их параметр независим от насадки. Единая команда посетителя условно координирует два механизма; это прямо поясняется в интерфейсе. Учебная модель использует доли 0,7 для действия рулей от скорости и 0,3 для действия насадки с тягой. Это выбранные коэффициенты упражнения, а не найденные характеристики Spiggen II. На ходу по инерции рули сохраняют действие; на неподвижной лодке без тяги команда глубины не действует. Балластные операции этим упражнением не воспроизводятся.