Resolve the fan
before running.

What the photographs show

Open the new fan close-up ↗ · Open the earlier close-up ↗ · Open the overhead context ↗ · Compare the earlier close-up ↗

Use the photo measurement worksheet to record a provisional visible gap if useful, then measure the clearance directly on the car. The photograph cannot establish a service tolerance.

Cold inspection on this car

1. Make the engine unable to start. Remove the key, let the engine cool completely, and keep hands clear of the belt path while documenting the fan.
2. Photograph every blade. Take a whole-fan image plus close views of each leading and trailing edge and the hub. Record blade count, molded markings, color, cracks, missing pieces, looseness and any rubbed area.
3. Record the mounting. Photograph the fan-to-coupling fasteners, coupling, water-pump nose and radiator clearance. Do not assume the photographed arrangement is unchanged today.
4. Check related visible faults. Look for coolant traces at the pump and radiator and obvious contact marks on the shroud. Record any looseness or leak as a separate finding; the tip photograph does not diagnose the clutch or pump.
5. Choose the branch. If the blade is intact, document why the edge is normal and keep other first-drive checks open. If any blade is damaged, plan a blade replacement. If the coupling or pump also has a confirmed fault, plan those as separate work and account for the tolerance ring when the coupling is removed.

Measure the fixing hexes

The CAD uses a GHF102 bolt/GHF200 nut catalog comparison with 7/16 inch AF hexes. That does not identify the hardware fitted here. Record opposite parallel flats separately at each numbered fixing; one reading must not stand in for all four. Use the same numbered photograph as the mounting-hole survey, and note head type, nut style and washer order. Record bolt hexes ↗ · Record nut hexes ↗. Height, thread, installed side and tool access remain unmeasured.

Rotation in the model

A fixed fan pose can miss a hose or cowl intrusion between blades. The public rotation audit checks the fan and its nominal bolts, nuts and washers with separate conservative bounds, then samples exact model contacts where needed. Each rotating solid is now checked against its bound by a native solid-volume difference; vertices alone cannot establish containment of curved faces. The report states its numerical containment tolerance and any residual outside volume. An undersized bound is rejected before an obstacle can be skipped. A blade-to-core gap alone does not bound the fixing projections. An overlapping envelope with no sampled contact stays inconclusive. It belongs to the default reconstruction, not a local owner build; run the source audit with the same owner log to inspect another build. These checks do not establish fitted gaps, running condition or a removal path. Actual blade profile, shaft pose and nearby surfaces remain unmeasured.

Record what you find

Fill this out beside the car. It saves in this browser and can be printed or downloaded as an inspection record. Keep the downloaded file with its original photographs; it contains photo filenames, not the images. Attach photos and repair costs in the project planner. Blank fields stay unknown. The download does not update the CAD model or establish replacement fit. The product photo suggests roughly 280 mm diameter and 67 mm adjacent-hole spacing. The current CAD has a 276.9 mm fixed X span and a distinct 279.0 mm swept diameter; neither is a measurement of this car. Leave the measurement fields blank until you measure this car. The public STEP remains the provisional study. To build a separate STEP from your measurements, select Mechanical fan in the CAD inspector, record the directly counted blades, maximum shaft-center-to-intact-tip radius and, when safely accessible, all four mounting-hole center positions with a source note, then export the measurement log. View the fan from the front of the car toward the engine; use the shaft center as zero, X positive to the right and Y positive upward. For a new record, number the holes clockwise from the uppermost one, choosing the left one if two are level. Preserve existing numbered photos and their coordinate records. A direct blade count changes repetition in a separate owner build but keeps the candidate paddle profile; a direct maximum tip radius controls the blade sweep. A whole-X reading controls only a reviewed fixed front-view span; old unspecified X readings remain notes. Only a complete set of four signed X/Y hole centers moves the shared fan, clutch and fixing pattern. One adjacent spacing remains a comparison note; it cannot establish a square pattern or identify the fitted fan.

Parts to compare before ordering

The catalog numbers below come from two distinct 1500 cooling arrangements. Rimmer lists RKC92 with the North American viscous coupling. Its Spitfire comparison plate lists the seven-blade RKC2126 with a direct, non-viscous water-pump pulley mounting. Match the actual fan, coupling, engine number, chassis number and hardware before purchasing; a changed cooling system is possible on a 1975 car.

The RKC92 product photo shows the open four-hole spider and blade shape used as a visual reference in the engine CAD. Compare it with the seven-blade RKC2126 photo to identify the fitted arrangement; blade count alone does not establish interchangeability. The CAD also shows a finned UKC3532 comparison coupling and four separate bolt and nut positions, with their catalog hardware sizes. The factory plate supports four fixings and depicts washers on the bolt-head side. The CAD now follows that catalog relationship and seats its estimated fan bosses against the coupling pads. Its axial stations, washer thickness and fastener heights remain estimates. Mounting positions, fitted direction and axial dimensions still need direct confirmation. Its proportions are not measurements of this car. Rimmer’s shared Spitfire plate (PDF p. 20) shows the same four-fastener fan-to-coupling relationship but lists a different plain-washer number. The Midget-specific listing is the better ordering starting point; the fitted hardware still decides.

Rebuilt fan and pump pulley in the main model

The existing IMG_6853–6855 views now guide the visible blade arc, with IMG_6895 showing the pronounced pitch. The old nearly flat fan and small pump pulley are replaced by a connected pitched fan and a hollow dished pulley. The pump, gasket, studs, coupling, ring and fan fixings retain their source dimensions on one shared construction axis. The original detailed engine export stays available as a historical comparison.

Inspect the revised main engine bay ↗ · Compare against the owner photograph ↗ · Download native front-drive STEP ↗

Evidence and assumptions: the shared photographic frame still depends on an unmeasured 152 mm cover-stud span. Three manually picked leading-tip proxies and reviewed reconstructed tracks guide a +Z shaft at estimated X/Y −58/−26 mm and a 30-degree blade phase. Thirteen equally spaced blades remain the RKC92 catalog comparison, not an identified fitted count. Tests with 12 and 14 blades also fit the partial arc and move the inferred center about 11 mm each way; the photos do not select the count. The selected 142 mm blade radial station, section profiles, 3.6 mm projected thickness and 40 mm axial shift remain construction estimates. Damaged edges are not reconstructed as a verified damage survey.

The visible dark rim/belt boundary lies about 100 conditional millimeters from the selected axis. It guides a 200 mm outside-diameter pulley comparison. Its 244 mm groove station, 13 mm width, V section, bowl, web and bore remain estimates; the visible boundary is not a measured groove center. The UKC774 supplier reference and factory pump/fan architecture support component roles, not those fitted dimensions.

What remains open: the revised static fan no longer intersects the upper hose or outlet source shapes. This is not a full rotational sweep or an operating-clearance finding. The pump receiver reference stations differ by 90 mm from the retained housing construction, and the nearest source surfaces remain about 31.8 mm apart. This is an explicit reconstruction gap, not a gap measured on the car. Original crank and photo-alternator groove comparisons differ from the new pump lane by 28.5 and 51.5 mm respectively. No belt, housing receiver or hidden coolant connection is invented to close those differences. The cover/hose source conflict from the preceding revision remains.

Workshop use: select the fan, coupling/ring, fixing stacks, pump and pulley separately to study their roles. Display separation is illustrative. Fitted identity, retention, ring compression, mounting compatibility, tool access, release sequence and replacement/removal path remain unverified. The following historical core audits describe their named source exports; they do not validate this new assembly. Read the construction evidence ↗ · Read native checks and open boundaries ↗

Historical core: ring and drive comparison

The native pump/ring/coupling joint audit ↗ finds the default reference ring detached from both neighboring solids: about 0.100 mm from the pump drive and 0.118 mm from the coupling socket at their nearest points. All three tested pairs have zero model overlap volume. These are discrepancies between estimated rigid shapes, not fitted operating clearances, a diagnosis of this car, or instructions to change a ring.

The ring manufacturer’s general explanation ↗ describes a tolerance ring as a radial spring joining mating cylindrical parts. That explains why merely drawing a sleeve between two components does not reconstruct the connection. It does not identify UKC759’s manufacturer, ring style, compression, holding force or installation specification.

During already-planned pump or coupling access, record the actual pump-nose crest/root profile, shoulder and projection, and the socket bore profile, depth and shoulder from named datums. Keep a new ring’s free-state dimensions and waves separate from the removed ring’s deformed condition; retain markings, original photographs and measuring method. Photograph the installed stack before separating it. None of these three interface profiles or their fitted axial stations is verified yet.

The source shapes remain unchanged. The joint report records the missing modeled contact rather than resizing independent estimates to force a fit. Even a future contact result would not establish interference, spring compression, retention, torque transfer or removal access. The separate rotation report omits relative pump/coupling motion; this static joint check does not fill that motion gap. The cooling viewer now preserves these source relationships while moving the fan, coupling and fixings together. Its forward offsets use the foremost displayed component and remain bounded to the comparison core plane even when that core is hidden. A 2 mm display margin is a viewing choice, not a service clearance; cowl, hose and tool access are not checked. For a local owner build, rerun this audit with the same measurement log instead of borrowing the public report.

Survey the bearing faces before reshaping the spider

One pad-thickness reading cannot establish the surrounding spider, mounting-pad outline, opposite face or installed offset. During already-planned access, use the blank fan bearing-face worksheet ↗ for direct spatial samples of one actual pad face. Set surveyed_component to fan or fan_coupling after identifying the part; keep the other component and the opposite face in separate records. The joint is fan_to_coupling.

Keep the pad tied to its numbered mounting-hole photograph. Record three ordered face points, independent face checks, directly observed closed hole or outer-edge traces, and independently measured chords between named samples. Do not infer circular holes, a missing outline or four coplanar pads. A caliper thickness or an oblique photo cannot supply these spatial coordinates. If the tool’s conditioning limits cannot be met on an accessible pad, retain the observations without fabricating more points.

The native survey and registration instructions ↗ explain how to retain the samples as separate reference chords and check vertices. A separately registered report can be opened in Measured joint references ↗ against its exact model file. An unchanged physical frame and independently checked CAD counterparts are required; a removed fan’s bench coordinates do not establish its fitted pose. The reference hides when fan, coupling, fixing or pump parts are separated, or a fan/pulley size preview is active.

This workflow does not resize the fan, calculate pad thickness, move the mating part, infer a contact surface or prove fit, clamping or removal clearance. It preserves the evidence needed to replace the estimated spider later. The existing manifold-named file format is shared by these joint-face references; this worksheet is not an engine measurement-log import.

Keep each fixing stack with its hole

When the fan is already off during planned work, keep each bolt, washer and nut indexed to the same numbered mounting-hole photo. Inspect whether there is one flat plain washer, a split or curved washer, a stack, or another arrangement. A catalog drawing does not establish what is fitted.

For one confirmed flat plain washer, measure between its two flat bearing faces along the bolt axis. Exclude burrs and do not substitute the height of a split washer or stacked pack. Record each thickness separately in the fan nuts and washers inspector ↗, with the numbered photo and measuring method. The owner builder accepts 0.3–4 mm for this plain-ring study after its form is confirmed. Outside that study range, retain the actual reading in the source note; do not change it to make the model accept it.

Each measured washer changes only its own thickness and moves its own bolt head and unchanged-length shank forward from the estimated fan face. The nut, coupling and fan mounting stations stay provisional. A thicker washer can leave a previously modeled shank short of the estimated nut interval; the model reports that mismatch and retains both readings. Even a covered interval does not prove threaded engagement, correct hardware, tightening or clearance.

Before matching a seller’s fan, ask for both faces, all markings, blade count, all four hole positions, the mounting-pad profile and thickness, and the hub’s front-to-rear offset. A matching blade diameter or one adjacent-hole distance is insufficient. The current CAD does not accept mounting-pad thickness as a fitted shape: its surrounding spider and bearing-face profile still need direct evidence. Washer measurements are not a substitute for those fan dimensions.

Before undoing the first fastener

The first geometry milestone is a shared installed layout of the shaft, radiator, cowl and hose connections. Follow the prioritized cooling-layout capture plan ↗ and use the replacement-readiness evidence table ↗ to track what still prevents a reliable rehearsal. Separate part dimensions do not locate the surrounding obstacles.

The radiator cowl is now a separate part in the 3D study. The owner overhead photo marks its visible bridge. Record clearance to that metalwork separately from clearance to the radiator core; the model cannot tell you whether the fitted blade will pass either obstruction.

Factory manual locator: the scanned contents page places 1975–76 USA cooling at manual page 324 and lists coolant drain/refill as 26.10.01, radiator removal as 26.40.01, water-pump removal as 26.50.01 and pump-housing removal as 26.50.03. Inspect the factory contents page ↗. It does not list a separate fan-blade operation or provide the actual removal instructions in this scan.

The accessible catalog plates identify assembly relationships and candidate parts, but they are not a full removal-and-refit procedure. The available photographed manual sections have not provided a complete, inspectable fan operation. This is the remaining source gap for a car-specific replacement guide. The original technical-publications catalog lists AKM3267 for North America 1975 and AKM4071B for the 1975–79 Midget 1500 ↗.

Source trail

Factory Sprite/Midget parts catalogue, PDF p. 64 ↗ lists the USA–Canada UKC3532 / RKC92 arrangement and four each of bolts, washers and nuts (items 26–28). The exploded drawing supports the catalog stack relationship, but supplies no fitted axial dimensions. Moss Sprite/Midget catalog, PDF p. 37 ↗ lists the 1500 fan, clutch, belt, pump and ring and explicitly requires a new ring when the clutch is removed and replaced. Rimmer’s North American Midget fan listing ↗ identifies RKC92, UKC3532, UKC759 and the fan fasteners. Rimmer’s Spitfire exploded plate, PDF p. 20 ↗ is a shared-engine comparison, not proof of this Midget’s fitted parts. The owner photographs establish the visible tip concern only. A separate owner’s Midget pump-repair account ↗ reports four fan fixing holes that were not equally spaced on that assembly; it does not identify this car’s fan or prove the same pattern here.

Start the fan project ↗ · Return to the cold cooling inspection ↗ · Identify the 1975 cooling configuration ↗

Historical core: cowl reconstruction conflict

The earlier upright 1500 cowl and sloping apron produced sampled material overlap with the historical source fan. The current whole-car front-drive construction is documented above. The updated rotation report ↗ retains this result; the fan and radiator were not repositioned to create clearance. It describes the current estimated source geometry, not measured contact on this car. Inspect cowl/body mounting references ↗. The separate factory CHA420 guard, fitted mounting coordinates and verified fan/removal clearances remain unresolved.