This case study turns the workholding lesson from the README into a reusable fixture-validation exercise. It is intentionally public and non-proprietary: it does not include employer-owned toolpaths, fixture CAD, post-processor settings, part drawings, dimensions, feeds, speeds, or confidential production details.
Use it with ../reference/fixture-validation-checklist.md
and ../skills/haas-cnc-workholding.md.
The public lesson is simple: a workholding failure taught that CNC setup validation starts before CAM review. The first question is what the part will try to do when the cutter touches it, and the second question is what holds it against that motion.
This case study is not a reconstruction of a proprietary job. It is a general training pattern for awkward near-net or low-volume CNC parts where datums, restraint, and first-cut observation matter more than elegant toolpaths.
| Field | Public-safe case-study value |
|---|---|
| Machine class | 3-axis VMC or CNC router equivalent |
| Part type | Irregular near-net, prototype, or one-off part |
| Stock condition | Not guaranteed to have finished datums |
| Fixture family | Vise, soft jaws, fixture plate, or spoilboard fixture |
| Authority level | Checklist training only |
| Excluded evidence | Proprietary drawings, CAM, toolpaths, fixture CAD, feeds, speeds |
The risky setup pattern is a part that appears secure under hand pressure but can still rotate, lift, chatter, or pull out under cutter load. That failure is easy to misdiagnose as a programming or feed-rate problem after the fact. The fixture-validation posture treats restraint as its own design problem:
- Name the likely cutter-force directions before reviewing the toolpath.
- Identify the datums that are trustworthy today, not the datums wished for on the final drawing.
- Decide how the part is prevented from pull-out, roll, lift, chatter, and jaw slip.
- Dry-run the full program path in the machine context.
- Inspect the part and fixture after the first meaningful engagement.
| Gate | Case-study question | Acceptance evidence |
|---|---|---|
| Stock registration | Is the setup referencing a real datum or a rough near-net surface? | Setup sheet names the datum, work offset, stock orientation, and worst-case envelope. |
| Hold-down / restraint | What motion does the cutter induce first: pull-out, roll, lift, push-off, chatter, or jaw slip? | Restraint strategy is named before CAM review, including secondary support for thin walls or long overhangs. |
| Spoilboard / sacrificial layer | Does the setup break through into a safe sacrificial layer, open air, or a vise/fixture hazard? | Spoilboard or N/A reason is recorded, and cutter exit paths avoid table, jaws, clamps, and fixtures. |
| Tool clearance | Can every tool, holder, and retract move clear clamps, jaws, walls, and sensors? | Dry-run or traced rapid path confirms tool length offsets and retract heights clear the tallest fixture feature. |
| Breakthrough | Could the cut free a part or slug while the cutter is still engaged? | Tabs, bridges, support, or a controlled release plan are present before cutting. |
| Dust / chip / coolant | Will chips, swarf, coolant, or dust degrade holding or visibility? | Evacuation method is named and the operator commits to clearing/inspecting before re-engagement. |
| Emergency stop / fault recovery | What happens if the setup moves or a stop is hit mid-cut? | Recovery plan names spindle/tool retraction, chip clearing, fixture inspection, and restart decision criteria. |
| Dry run | Has the actual machine motion been checked outside the CAM preview? | Program is cycled with spindle off or safe Z/retracted state, then first real cut uses reduced feed observation. |
Use this as a public case-study review checklist before taking a fixture plan to an operator, mentor, or shop reviewer for first-cut approval:
- The setup sheet names the machine class, part, stock condition, fixture, operator, reviewer, and work offset.
- The primary restraint strategy is named before toolpath review.
- The force directions most likely to move the part are listed.
- Each clamp, screw, jaw, stop, tab, bridge, and support is outside cutter and retract paths.
- Any
N/Agate includes a reason tied to the machine class. - CAM preview is not treated as machine validation; an air run or safe dry run is still required.
- The first material engagement has a reduced-feed observation plan.
- The operator inspects the fixture after the first meaningful cut.
- No proprietary geometry, toolpath, feed, speed, post, or fixture design is required to understand the lesson.
This repository does not contain measured validation for a specific fixture. When a real public fixture is available, record measured evidence separately:
| Record | Example public field | Why it matters |
|---|---|---|
setup_id |
Public fixture or training setup ID | Keeps observations traceable. |
gate_id |
Stock registration, hold-down, dry run, etc. | Links evidence to the checklist. |
target |
"No jaw slip after first engagement" | States the intended behavior. |
observed |
Pass/fail/pending/N/A plus note | Separates evidence from aspiration. |
measurement |
Deflection, witness mark, torque check, or dimensional delta | Turns the story into a learning loop. |
follow_up |
Re-cut soft jaws, add stop, revise tabs, rerun dry pass | Makes the next setup better. |
Before publishing any follow-on case:
- Remove customer, employer, project, material-lot, drawing, and part-number details unless explicitly cleared for publication.
- Do not publish G-code, CAM screenshots, post-processor settings, fixture CAD, or proprietary dimensions.
- Keep photos cropped to show the validation principle, not confidential part geometry.
- Label aggregate or redacted observations as such.
- Keep readiness language at "case-study checklist" until measured setup evidence exists.
This case study advances the public documentation but does not close the whole fixture-validation problem. Remaining gates:
- Apply the checklist to a real public fixture or training setup.
- Record pass/fail/N/A evidence for each gate.
- Add measured observations after the first cut, not only narrative lessons.
- Cross-link the resulting measurement record without exposing proprietary CAM, drawings, fixture designs, or production details.