Zipper Teeth Pitch Control: Diagnosis and Process Capability

Symptom

Pitch deviation, the centre-to-centre distance between adjacent teeth, is the most frequent dimensional complaint on a finished zipper. In a defective lot it rarely appears as random scatter. It appears as a repeating signature: the slider runs freely for 60 to 80 mm, hesitates, releases, then hesitates again. In garment service the same lot produces intermittent chain opening at the chest or seat line, and the complaint rate climbs after the first domestic wash cycle.

Because pitch error accumulates, a 0.05 mm error per pitch becomes 5 mm over one hundred pitches. That is enough to pull a nominal 600 mm zip outside its length tolerance, to skew the tooth profile so the interlocking geometry no longer self-centres, and to overload the slider body on one side. Inspectors therefore see two symptoms at once: a measured length that drifts from the drawing, and a hand-feel that changes along the chain.

Typical measurable indicators include:

• Pitch measured over 10 teeth varying by more than 0.08 mm within a single metre.
• Finished length outside ±0.5 percent of nominal on a 500 mm sample.
• Slider pull force rising above 12 N on a 5# nylon chain at the tight section.
• Visible tooth crowding at one end of the chain and a gap at the other.
• Chain split reported at a repeatable height on the garment, not randomly.

Root Causes

Pitch is created by the interaction of four variables: mould or die geometry, chain advance, cooling, and tension. Any drift in one of them modulates the others, which is why pitch problems so often look like several faults at once.

• Worn mould cavity or die insert. Cavity spacing opens by 0.02 to 0.04 mm after roughly two million shots; the error is uniform across the cavity and therefore shows up as a slow, whole-lot drift rather than a spike.
• Chain advance error. A slipping servo coupling, a contaminated encoder wheel or a stretched timing belt produces periodic error synchronised with the drive revolution. This is the classic "smooth then tight" signature.
• Uneven cooling. Nylon monofilament that is still soft when it enters the next station stretches under chain tension, so pitch grows progressively along the run. The first metre is good and the last metre is long.
• Tension drift on the winder. Take-up tension that rises through the shift compresses pitch at the end of each package, creating a characteristic within-package gradient.
• Incorrect monofilament diameter. Oversized filament cannot seat fully in the cavity and pushes neighbouring teeth apart; undersized filament sits loose and reads short after setting.
• Post-process heat. Dyeing or coating above the polymer's dimensional stability threshold relaxes the chain and changes pitch after the part has already passed dimensional inspection.

Troubleshooting Sequence

Work from the finished part backwards. Do not adjust the machine until the error signature is characterised, because a random adjustment on a periodic fault makes the next sample unreadable.

• Sample three positions per package — start, middle, end — and measure pitch over ten teeth at each position with a calibrated optical comparator or a pitch gauge held at 20 ±2 °C.
• Plot the three positions against machine direction. A rising trend indicates cooling or tension; a flat offset indicates tooling.
• Compare two consecutive cavities or two adjacent lines. If only one is affected, the fault is local to that tool or that drive.
• Inspect the mould or die for flash, wear marks and polishing damage on the cavity parting faces.
• Log the servo following error and encoder output for one full shift, looking for a period that matches the observed pitch cycle.
• Measure incoming monofilament diameter on the same spools used for the suspect lot; retain the certificate.
• Re-measure after dyeing and after coating to separate pre-process from post-process contribution.

Corrective Actions

When the error is tooling-related, replace or refurbish the cavity insert and re-qualify with a 30-piece first article. When the error is periodic, replace the encoder coupling or belt and confirm that the pitch cycle disappears over three consecutive packages. When cooling is the driver, reduce line speed by 8 to 12 percent or extend the cooling tunnel, then verify that the start-to-end gradient falls below 0.05 mm.

For accumulated length error, do not simply stretch or shrink the finished chain. Recalculate the cut length from the measured pitch, adjust the cut-to-length offset, and re-validate the top stop position, because moving the cut point without moving the stop station transfers the error into the stop location. Document the new offset in the machine recipe so that the correction survives the next changeover.

Prevention

Pitch control is a capability problem, not an inspection problem. Establish a control chart on pitch measured every two hours, with limits derived from the actual process spread rather than from the drawing tolerance alone. A process running at Cpk 1.33 on a ±0.10 mm band gives roughly one escape per million teeth; a process at Cpk 0.9 will generate complaints even though every individual part passes a go/no-go gauge.

Add three structural safeguards. First, schedule cavity wear checks by shot count, not by calendar. Second, keep a golden sample of 200 mm chain at 20 °C in a sealed tube and verify the pitch gauge against it at the start of every shift. Third, require that any change to dye temperature, coating line speed or winder tension triggers a pitch re-validation before release, because those three parameters can invalidate a dimensional result that was correct at the time it was taken.

Donglong Zipper applies this sequence on every forming line, and YCC auditing teams review the pitch control chart together with the mould shot counter during supplier assessments. When both records agree, pitch complaints typically fall below 0.3 percent of shipped cartons within one quarter.