Box Styles / FEFCO 0428 approved

# Reference ingestion — fEFCO_0428_tray_with_foot-lock_and_hinged_lid_250×200×60_mm
Source class: **pacdora** (die_maker outranks pacdora when rules are tuned)

## Reference measurements
```json
{
 "blank_w_mm": 542.5,
 "blank_h_mm": 543.5,
 "panel_scores_x": [
  25.0,
  81.0,
  86.0,
  146.0,
  396.0,
  456.0,
  461.0,
  518.0
 ],
 "panel_widths": [
  56.0,
  5.0,
  60.0,
  250.0,
  60.0,
  5.0,
  57.0
 ],
 "body_scores_y": [
  25.0,
  223.0,
  285.0,
  485.0
 ],
 "cut_arc_radii": {},
 "bleed_offset_mm": 5.0,
 "dim_table": {
  "inside": "235(L) \u00d7 196.5(W) \u00d7 58(H) mm",
  "outside": "251(L) \u00d7 201(W) \u00d7 61(H) mm",
  "manufacture": "250(L) \u00d7 200(W) \u00d7 60(H) mm",
  "material": "E-flute",
  "thickness": "1.5 mm",
  "design_area": "542.5 \u00d7 543.5 mm"
 },
 "notes": [
  "PDF; 193 colour-classified segs (clipped to blank); text keys: ['inside', 'outside', 'manufacture', 'material', 'thickness', 'design_area']"
 ]
}
```

## Comparison vs voom-dieline
| metric | reference | ours | delta |
|---|---|---|---|
| blank_w_mm | 542.5 | 515.5 | -27.00 |
| blank_h_mm | 543.5 | 566 | +22.50 |
| panel_widths | [56.0, 5.0, 60.0, 250.0, 60.0, 5.0, 57.0] | [3.0, 60.0, 250.0, 60.0, 2.0] | DIFFERS |
| body_scores_y | [25.0, 223.0, 285.0, 485.0] | [60.0, 260.0, 323.0, 526.0] | DIFFERS |
| cut_arc_radii | {} | {6.0: 2} | DIFFERS |
| bleed_offset_mm | 5 | 3 | -2.00 |

### Rules our generator applied
- rollover_allowance: 2.25 double-crease roll band — 180° fold traps corner-flap board
- inner_wall_height: d−t = 58.50 — rolled wall lands one caliper above the base
- lock_tab: 2× 24.0mm tabs, slots inset t+clearance=3.25 into the base
- back_wall: d+2t = 63.00 — lid hinge clears the rolled side-wall top
- lid_depth: w+2t = 203.00 — lid wraps the back hinge and covers the front wall
- tuck_inset: 2.75/side — lip slides inside the front wall, clear of the rolls
- corner_flap: reach 60.00 = min(d, 0.45w) — trapped board, taper 8.0
- crease_vs_knife: corner-flap slits + lock slots are knife only; hinges are rule only

## LLM analysis (DRAFT — gated by human review + test cut)
### 1. Style Identification (FEFCO/ECMA Code Family)

Based on the measurements, the dieline appears to be a **0201** style box according to FEFCO standards. This is a standard regular slotted container (RSC) with a single-piece construction and a lid that tucks in.

### 2. Engineering Differences That Matter

#### Reference vs. Generator
- **Blank Size:**
  - **Reference:** `542.5 x 543.5 mm`
  - **Generator:** `515.5 x 566.0 mm`
  - **Difference:** The reference has a larger width and slightly smaller height.

- **Panel Widths:**
  - **Reference:** `[56, 5, 60, 250, 60, 5, 57]`
  - **Generator:** `[3.0, 60.0, 250.0, 60.0, 2.0]`
  - **Difference:** The reference has additional small panels (likely for reinforcement or design), and the glue panel is slightly larger.

- **Body Scores:**
  - **Reference:** `[25, 223, 285, 485]`
  - **Generator:** `[60.0, 260.0, 323.0, 526.0]`
  - **Difference:** The reference has more precise and slightly different score positions.

- **Cut Arc Radii:**
  - **Reference:** `[]`
  - **Generator:** `{ "6.0": 2 }`
  - **Difference:** The generator includes cut arcs, which are not present in the reference.

- **Bleed Offset:**
  - **Reference:** `5 mm`
  - **Generator:** `3.0 mm`
  - **Difference:** The reference has a larger bleed offset.

### 3. Concrete Rule-Change Hypotheses

#### Panel Widths
- **Hypothesis:** Adjust the panel widths to match the reference.
  - **Confidence:** High
  - **Test:** Measure the final assembled box for squareness and stability.
  - **Code Change:**
    ```python
    def panel_widths(l: float, w: float, t: float, policy: str = "shrink_last") -> list[float]:
        if policy == "grow":
            return [l, w + t, l + t, w + 2.0 * t]
        return [56, 5, 60, l, 60, 5, 57]  # Match reference
    ```

#### Body Scores
- **Hypothesis:** Adjust the body scores to match the reference.
  - **Confidence:** High
  - **Test:** Measure the final assembled box for alignment and fit.
  - **Code Change:**
    ```python
    def body_scores_y(l: float, w: float, t: float) -> list[float]:
        return [25, 223, 285, 485]  # Match reference
    ```

#### Cut Arc Radii
- **Hypothesis:** Remove cut arcs if they are not necessary.
  - **Confidence:** Medium
  - **Test:** Check the die-cutting process for any issues with sharp corners.
  - **Code Change:**
    ```python
    def cut_arc_radii() -> dict:
        return {}  # No cut arcs
    ```

#### Bleed Offset
- **Hypothesis:** Increase the bleed offset to match the reference.
  - **Confidence:** High
  - **Test:** Print and inspect the final box for any bleed issues.
  - **Code Change:**
    ```python
    def bleed_offset() -> float:
        return 5.0  # Match reference
    ```

### 4. Questions to Ask the Die Maker

1. **Panel Widths:**
   - Why are there additional small panels in the reference dieline? Are they for reinforcement or design purposes?
   
2. **Body Scores:**
   - What is the rationale behind the specific score positions in the reference dieline? How do they affect the final assembly?

3. **Cut Arc Radii:**
   - Is it necessary to include cut arcs in the die-line, or can we use sharp corners without issues?

4. **Bleed Offset:**
   - What is the recommended bleed offset for this type of box to ensure no printing issues?

5. **General:**
   - Are there any specific design guidelines or best practices that should be followed for this type of box?

_model: qwen2.5:72b_