Background
A European premium brand commissioned a custom slider for a limited-edition jacket. The slider was a #5 auto-lock body with a raised logo on the pull face and a brushed finish. Because the logo had fine strokes and the part was small, the project was treated as a tooling development rather than a standard catalog purchase. The order quantity was modest, around 12,000 sliders, but the visual standard was high because the slider sat at eye level on the chest.
The brand supplied a 3D model and a rendering, and the supplier quoted a two-cavity mold with a four-week tooling lead time. The first samples were promised before the garment pre-production meeting so that the brand could photograph the jacket for its campaign.
What Went Wrong
The first tooling trial produced sliders with incomplete fill at the logo edges and a visible weld line across the pull face. The second attempt improved fill by raising injection pressure, but the logo strokes became rounded and the part flashed at the parting line. The third attempt changed the gate location, which fixed the flash but introduced a sink mark on the opposite face and made the logo shallower still. By this point the pre-production meeting had passed without an approved slider.
The brand began to question whether the design was manufacturable at all, and the supplier began to question whether the brand's model was dimensionally sound. Neither side had a shared measurement of what "good" meant, so each trial was judged by opinion rather than by a drawing with tolerances.
- Trial 1: short shots at logo edges, weld line on pull face.
- Trial 2: acceptable fill, rounded strokes, flash at parting line.
- Trial 3: flash resolved, sink mark and shallow logo.
- Status: pre-production date missed, no approved part.
Root Cause
Mold flow analysis, run only after the third failure, showed that the gate was feeding the thickest section last, so the thin logo area froze before the cavity filled. The original gate position had been chosen for ease of machining rather than for fill balance. Shrinkage was also unaccounted for: the supplier applied a generic shrinkage factor to a glass-filled polymer, which pulled the logo depth below the minimum for legibility.
A second issue was specification. The brand's model defined the logo as a surface texture with no depth tolerance, no draft angle and no minimum stroke width. The toolmaker interpreted it literally, which produced a geometry that could not be molded reliably at that scale. The parting line had never been agreed either, so the flash appeared in a visible location.
| Trial | Main Defect | Root Cause |
|---|---|---|
| 1 | Short shot, weld line | Gate fed thin section last |
| 2 | Rounded logo, flash | Pressure raised, parting line unset |
| 3 | Sink mark, shallow logo | Gate moved, shrinkage misapplied |
How It Was Resolved
The brand, supplier and toolmaker held a technical review with a mold-flow engineer. They agreed a revised part drawing with a logo depth of 0.15 mm minimum, a 2-degree draft on all vertical faces, a minimum stroke width of 0.35 mm, and a defined parting line on the underside of the slider body. The gate was relocated to feed the thick body section first, with a overflow well at the far end of the logo to vent the last-filled area.
The toolmaker cut a fourth cavity insert rather than modifying the whole mold, which limited cost and lead time. Trial four produced full fill, a clean logo and no flash in the visible zone. The brand approved cosmetic samples under a defined viewing distance and light source, and the supplier added a first-article inspection report with a cavity-by-cavity photo record.
For the production run, the supplier added a 100% visual check on the logo face and a sampling check on pull force. The mold was registered in a maintenance system with a shot counter and a 50,000-shot service interval. The brand received its campaign photography two weeks later than planned but before the retail deadline.
The insert approach proved important for cost control. Because the logo cavity sits in a replaceable insert, the next cosmetic revision required only a small insert change rather than a full mold rebuild. The brand later added a second logo variant for a collaboration and paid for one insert, which kept the tooling budget predictable.
The maintenance record changed the supplier's behavior as well. With a shot counter and a 50,000-shot service interval, the mold is pulled for inspection before wear appears in the part. At the first service the toolmaker found polishing wear on the gate that would have produced a visible mark within another 20,000 shots. Catching it early cost one day of maintenance; missing it would have cost a production lot and a second cosmetic approval.
Lessons Learned
- Run mold flow before cutting steel, not after the third failure.
- Define logo depth, draft and minimum stroke width on the drawing with tolerances.
- Agree the parting line and the cosmetic viewing standard in writing.
- Use insert tooling for cosmetic features so revisions stay cheap.
- Register custom molds with a shot counter and a service interval from day one.
Custom sliders fail for ordinary reasons: fill balance, shrinkage and an under-specified drawing. None of the three trials was a matter of bad luck. Once the team replaced opinion with a measured part definition and a flow simulation, the fourth trial passed on the first attempt. Donglong Zipper now requires a signed cosmetic specification before any custom slider mold is released to steel.