The five saved gear images are anonymous user-supplied references. Their dimensions are transcribed data, not manufacturer certification. No price or manufacturing-tolerance claim is imported.
The four presets use module 0.5, 11 / 13 / 15 / 17 teeth, and a 2.98 mm bore, as shown in public/references/spur-pinion-options.png. Width 5 mm, pressure angle 20°, and tooth-thinning allowance 0.02 mm are editable prototype assumptions. The image does not specify these values.
The new small-pinion layout option permits tooth counts below the unshifted rack-cutter undercut limit. It retains sampled involute flanks with radial root relief; it does not imply a manufactured profile shift or reproduce cutter undercut. The existing unshifted mode continues to reject those low tooth counts.
The six presets contain both members of every pair in public/references/bevel-gear-dimensions.png:
| Module | Pinion / wheel | Pinion bore interval | Wheel bore interval |
|---|---|---|---|
| 1 | 20 / 40 | 6–8 mm | 8–12 mm |
| 1.5 | 15 / 30 | 6–10 mm | 8–12 mm |
| 1.5 | 18 / 36 | 8–12 mm | 10–16 mm |
| 1.5 | 20 / 40 | 8–12 mm | 12–20 mm |
| 2 | 15 / 30 | 8–15 mm | 14–20 mm |
| 2 | 20 / 40 | 12–16 mm | 16–30 mm |
All twelve table rows are retained in src/catalog/reference-gears.ts. The configurator exposes D outside diameter, A back-face mounting distance, F overall axial length, G large-end tooth-tip position, Z slant face width, N hub diameter, H hub / large-root position, and L small-end root position. Axial positions are measured from the back hub face.
The assembly origin is the intersection of the perpendicular shaft axes. Assembly and separated views export two independent solids. Pinion-only and wheel-only views center the selected part on its own bore axis. The six mounting references allow continuous bore adjustment within the supplied intervals. Discrete stock options are recorded separately from the additional m2 listing below.
Preset search compares requested bore values or ranges with those source intervals. For example, a 7 mm pinion bore finds the 6–8 mm reference. Results distinguish the source interval from the preset's selected bore; applying the preset retains the values shown in its review.
Both models include editable smooth radial set screw holes. Their diameter and thread detail are unspecified by the image, so the default 3 mm hole is not a source-verified parameter. Pressure angle and tooth thinning are also prototype settings.
The paired teeth retain the supplied envelopes with a faceted, tapered involute-like profile. They are not generated conjugate bevel tooth surfaces. The source has insufficient information for tooth corrections, root fillets, contact analysis, strength ratings or interchangeable manufacturing geometry. Linked rotation illustrates the tooth ratio; it does not certify transmission contact.
Both root annuli are planar at H and L. Local radial patches connect these planes to the raised tooth-end contours at G and F. Full flank samples retain the involute curvature; the module does not triangulate across unrelated teeth or stretch a tooth-tip height through the central face. The preview and FreeCAD share this boundary construction.
The additional mounting table, pair photo and bore-options image add 25 configurations, giving 31 bevel-pair presets in total. The listing describes separately sold gears made from 45# steel. Pinion bores are 8, 10, 12, 14 and 15 mm, all round. Wheel bores are 14, 15 and 16 mm with a 5 mm keyway, or 18 and 20 mm with a 6 mm keyway. The selected pinion and wheel bores have separate catalog selectors.
Stock presets match these exact bore values. The keyway depth defaults to 2 mm as an editable prototype value; the supplied image specifies only its width. Changing that depth does not turn it into a source-verified dimension. The original six mounting references retain their continuous bore intervals.
The end-face repair has a 17-case native FreeCAD audit (generated local report). It covers all six catalog pairs, all four display states on the first two pairs, hex/D/keyway bores, maximum bores at 37° rotation, and 128 inside/outside root-plane probes. All checked assemblies have zero measured component intersection. Every case preserves valid separate solids through STEP and FCStd round trips. Preview tests additionally sample both planar annuli on both gears for every preset and check closed, consistently oriented triangles for all seven shaft profiles.
To reproduce a small native smoke check, run node --import tsx scripts/verify-bevel-pair.ts, then run scripts/verify-bevel-pair.py --intersections using a FreeCAD-enabled Python. The preparer's optional --matrix expands to every catalog state and all seven bore shapes.
Use --stock-bores instead of --matrix to prepare five assembled pairs covering every listed pinion and wheel bore. This batch additionally probes the round openings and keyway walls at three axial positions, then checks STEP and FCStd round trips.
The stock-bore native audit (generated local report) passed all five pairs in FreeCAD 1.0.2, including 150 bore/keyway probes and independent component movement. All measured pair intersections were zero. Maximum preview/native differences were 0.0475% in volume and 0.000409 mm in bounds.
The reference expansion was checked in native FreeCAD using 46 cases: all assembly and separated states, both bore endpoints for each individual bevel gear, linked 37° rotation, and the four miniature spur pinions. All generated solids and STEP round trips were valid. The maximum preview/native volume difference was 0.067%; all checked pair positions had zero component intersection. This sampled geometry check does not establish continuous conjugate contact or a load rating.
A separate 48-case preview check covered six bevel pairs, both bore limits, two rotations and both independent component exports. All meshes were closed and outward oriented. The automated gear test suites also retain the existing involute, helical, rack and generic bevel regressions.