[family] stop_bar_tailpiece: Gibson-style stop tailpiece (replaces #10 example) - #79
Open
BenchCAD wants to merge 10 commits into
Open
[family] stop_bar_tailpiece: Gibson-style stop tailpiece (replaces #10 example)#79BenchCAD wants to merge 10 commits into
BenchCAD wants to merge 10 commits into
Conversation
Gotoh GE101Z class, replacing drum_tension_lug (#10 -> #78): crowned bar with six string holes (stepped bores feature), full-length base plate with open U-slot stud ears. 10 params. bench2 validate passes; renders inspected against the official Gotoh 2D drawing. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
bench2 validatestop_bar_tailpiecePreviews: the |
Codex (gpt-5.6-sol): the transverse section stayed rectangular while the GE101Z end view is a flat-backed D. Constructed fix — intersect with a front cylinder along the bar (crest at the front face); edge fillets failed BRep at hard extremes, the constructed intersect validates 15/15 (the per-body gate caught the fillet failure). Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
Collaborator
|
Should address the curved surface |
…ders -> ears) Maintainer review: the body-to-ear transition was a hard step, not the real part's flowing surface. Rebuilt as ONE ruled loft of constant-topology D-sections (~25 stations: crown arc over the body, cosine blend down each shoulder, flat ears), stadium plan intersect for the rounded ear horns, cuts after. ruled=True + dense stations because smooth lofting overshoots (bbox x+6.4, z+5) in the pinned OCC and the overshot surface silently empties later booleans; edge fillets for the D-profile die at hard extremes (constructed arc sections instead). New param ramp_len (proportion); blend/crown checks reworked to match the real outer-bore placement. validate 15/15 PASS. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
Per the GE101Z sheet: bores sit high so the phi5.1 counterbores break out through the crown top (the drawing's top view shows all six as open notches), and the stepped feature is now counterbored from BOTH faces (4.5-deep class) with the phi3 web, not a single-sided step. Pocket-floor check replaces the old mid-body hole-fit check. validate 15/15 PASS. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
… review) Independent image-only review vs the GE101Z sheet caught the one real geometry error: the plan view measures the 8 mm slot ALONG the bar and the slots open through the FRONT long face with closed, curled hook tips — not outward through the ends (the model had two-pronged forks). Reoriented the slots (front-opening, semicircular closed end past the stud centre), added the rounded ear lobes the plan view shows around each stud, and reworked the closed-tip constraint. validate 15/15 PASS. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
…dex max) The end view reads flat-BOTTOM / domed top: 12.75 is the HEIGHT and 18 the front-back depth (the long phi-3 waist between the 4.5+4.5 counterbores only fits through 18) — the previous model stood the section on end. Reworked on that basis: low loaf with R300 crown along the domed top, string bores low in the tab band (elevation), double 5.1 counterbores with the long 3 mm web, plan-view ear lobes clipped to the faces with the 8 mm U-slot leaving the curled outboard tip. Ranges/checks follow the corrected axes. validate 15/15. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
…surface quality) Near-coincident booleans were leaving sliver seams: the ear/body union overlapped by only 0.05 mm (now 1 mm, buried), and the slot's tangent cylinder+box cutters were cut sequentially (now fused into one cutter per side so the tangent seam stays in the cutter). Guarded the ear-lobe sqrt at extremes. Smooth lofting was re-tested with dense stations and still collapses in the pinned OCC — ruled stays. validate 15/15 PASS. Previews regenerated with the point-merged renderer (see the framework PR): station seams no longer draw, only true feature edges. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
Maintainer spotted tiny protruding faces at the body/ear junctions. Probes confirmed two defects: the rectangular ear section poked 0.4 mm3 past the loft's D-curve at the join, and the blend tail stood proud of the flat ear top (1.5 mm3 wedge). Fixes: ears now carry the same D-section (hair-inset 0.06/0.03 so the overlap nests strictly inside the loft instead of surface-on-surface), the ear plan clips 0.15 inside the faces (kills a knife-edge tangency at the D apex that broke tessellation), the loft runs flat 1.4 past the join, and the ramp uses 14 stations. Previews rendered with the point-merged renderer (#111). validate 15/15 PASS; junction probes now 0.00 mm3. Signed-off-by: BenchCAD <benchcad@users.noreply.github.com>
…family/stop_bar_tailpiece
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Family
stop_bar_tailpiece—cat:music-hardware— replaces [family] drum_tension_lug #10's example (see [family] stop_bar_tailpiece #78)Reference (issue #78 — Gotoh GE101Z official 2D drawing)
Generated — benchmark views (what the model sees)
Three-view + iso — hard example
Difficulty grid (easy/medium/hard x seeds)
Extremes — smallest & largest sampled instances
Surface construction (maintainer review: model the real blended form)
The whole outer form is one ruled loft of constant-topology D-sections (flat
back + front arc) at ~25 stations: height follows the R-crown over the body,
then a cosine blend down each shoulder into the ears; a stadium plan intersect
rounds the ear horns around the open slots. Engineering notes: smooth
(
ruled=False) lofting overshoots in the pinned OCC (bbox grew +6 mm and theovershot spline surface silently emptied later booleans) -> ruled + dense
stations; edge fillets for the D-profile fail
BRep_APIat hard extremes ->the profile is constructed in the section wire instead.
ramp_lenis the newblend-length parameter. String bores follow the sheet exactly: set high so the
counterbores break out through the crown top (the top view's six notches), and
the stepped feature counterbores from BOTH faces with the φ3 web.
How it's parameterized
Bar only (studs/bushings are separate hardware, split noted in family.json):
crowned centre body (R250-R300) with six string holes — stepped φ5.1→φ3 bores as
the
stepped_holesfeature — over a full-length plate whose open U-slot earshang on the two studs. 10 params; stud span pinned to the 82/82.55 standard.
Checklist
bench2 validate --seeds 5PASS; all four renders committed and compared against the official 2D drawingCloses #78; DCO-signed; sources cited (Gotoh drawings + 4 vendors)Sources used