Herringbone Annulus Gear

gears/ring/herringbone_annulus_gear.py → OBJECT_OT_herringbone_annulus_gear (object.herringbone_annulus_gear, "Herringbone Annulus Gear")

Herringbone internal gear — V-shaped involute teeth cut into the bore of a solid ring. See README.md for family-wide conventions and annulus_gear.md for the swapped addendum/dedendum geometry shared by all three annulus generators.

Hand convention

Same rule as the helical annulus gear: a herringbone annulus and its mating herringbone pinion need the SAME hand (unlike external-external herringbone pairs, which need opposite hands). Picking a target here runs sync_helical_same. See helical_annulus_gear.md for why this matters.

Properties

Note the property labels here ("Total Width", "Slices per Half") are more explicit than the plain "Width"/"Slices" labels on the helical annulus gear, because herringbone splits both properties' semantics across two mirrored halves.

Build method

No boolean, no Solidify modifier — same direct-bmesh rewrite as the straight and helical annulus gears (annulus_gear.md, helical_annulus_gear.md), extended to a V-shaped (herringbone) twist. This used to be a solid outer cylinder minus a boolean-DIFFERENCE herringbone cutter (EXACT solver).

_build_herringbone_annulus_solid builds:

• The inner (toothed) wall: 2*n_slices - 1 V-twisted Z-layers of the tooth profile — bottom half twist rising 0 → peak_twist over [0, width_mm/2], top half falling peak_twist → 0 over [width_mm/2, width_mm], sharing the mid-slice at the peak (the top-half loop starts at k=1 to avoid rebuilding that shared slice). Twist is relative to z=0/width_mm directly now — the old boolean cutter had to pad its Z range by ±BOOL_EPSILON to avoid a coincident-cap boolean artifact, which no longer applies since there's no boolean. peak_twist = (width_mm/2) * tan(helix_angle) / pitch_radius.
• The outer (cylindrical) wall: a PLAIN, UNTWISTED, independently- spaced circle (outer_segs points, only 2 Z-layers) — the outer surface doesn't twist, only the inner teeth do. As with the other two annulus generators, this is deliberately NOT built with one point per tooth-profile point at a matching angle (that reintroduces zero-area triangles at the tooth profile's collinear dedendum-circle point insertion).
• The two end caps: bmesh.ops.triangle_fill fed the boundary edges of both the (twisted) inner ring and the (untwisted) outer ring together at each end.

Pressure angle clamp has extra margin here too: _derived() clamps to gear_matching.max_pressure_angle_deg(...) - PA_TRIANGLE_FILL_MARGIN_DEG (0.2°), not the theoretical limit itself — triangle_fill is more fragile right at that boundary than the old EXACT-solver boolean was. Swept tooth counts 8-100 × pressure angles 10-45° × both hands: 96/96 clean.

herringbone_planetary_gear_set.py's own ring gear is unaffected by this rewrite — it's a separate file with its own cutter function (_make_herringbone_cutter_obj), still boolean-based.

Panel warnings

Only tip_r <= 0 → "Module too large — tip radius ≤ 0" (ERROR, blocks). Info box additionally shows normal module, peak twist, and total slice count (2*n_slices - 1).

No print-in-place clearance logic — produces a rigid single-part ring meant to be printed separately from its pinion.

Output

One object, stamped:

gear_matching.stamp_gear(body, "herringbone_annulus", module, pressure_angle_deg,
                          tooth_count=tooth_count,
                          helix_angle_deg=helix_angle_deg, hand=hand)

Meshes with a herringbone pinion (herringbone_gear.md) of the same module, pressure angle, and helix angle, and the same hand.