Quick answer: A liquid oral dose line was producing reliable seals only below about two thirds of its rated line speed. Two liner suppliers and one sealer service visit had failed to fix it. The cause was a liner diameter specified from the closure catalogue rather than from the moulded neck: the effective sealing land was narrower than the nominal finish implied, so the sealant was being asked to bridge an overhang. Correcting the diameter restored full line speed with no change to the sealer.
This note describes a project at Hangzhou Xinxing Pharmaceutical Packaging. It is published because the failure pattern is common and the diagnostic sequence is short, and because the cost of the problem — in this case, months of reduced output — was far larger than the cost of the fix.
The situation
A manufacturer of liquid oral dose products in HDPE bottles was running a filling line with an induction sealer positioned downstream of the capper. At the line's rated speed, seal integrity tests showed failures at a rate that the quality team would not release. Reducing speed to roughly two thirds of rated eliminated the failures, and the line had been running there for some months.
The liner had been replaced once with a different supplier, and the induction sealer had been serviced and retuned. Neither changed the outcome. By the time the enquiry reached us, the working assumption inside the plant was that the liner specification was fundamentally wrong and that a different construction was needed.
What we asked for first
Three things, before any discussion of materials:
The actual production containers, not samples from a design studio or a pilot tool.
A video of containers passing the sealing head at the speed that failed, close enough to see the coil gap.
The liner drawing and the diameter actually being supplied.
What the measurements showed
The liner was specified at the diameter implied by the neck finish — the number a closure catalogue gives for the relevant thread designation. Measuring the moulded container at three points around the rim showed an effective sealing land narrower than that diameter by a margin that varied around the circumference.
The practical consequence was an overhang. The liner disc extended past the sealing land enough that, when the cap was torqued down, the sealant film bridged the gap between the land and the outer edge of the liner rather than being pressed flat against the land. At slow speed, with a longer dwell, the field delivered enough energy for the bridging sealant to wet and bond. At rated speed the dwell shortened, the bond thinned, and containers began to fail.
This explains why both previous interventions missed. A different liner construction of the same diameter kept the same overhang. Retuning the sealer increased energy, which masked the problem at moderate speed and would have pushed the sealant past its upper limit at full speed.
The fix
A new die at the measured sealing land diameter, with the diameter confirmed against a second set of containers from a different mould cavity. Construction otherwise unchanged — the same sealant polymer, the same foil thickness, the same board backing, because none of those had been shown to be at fault.
The line was re-qualified with a power sweep at fixed speed, then a speed sweep at the chosen power. Rated speed was recovered with margin at both bounds. No change was made to the sealer settings that had been in place before the problem started, which is worth noting: the sealer had been adjusted upward during the investigation and was returned to its original setting.
What the case illustrates
A nominal finish is not a sealing dimension. The finish designation describes a thread and a closure. The sealing land is a moulded feature and it has to be measured on a production part. The gap between the two is frequently a fraction of a millimetre and occasionally much more, and the liner diameter has to match the second of them.
An overhanging liner fails in a way that looks like under-sealing. The visual signature — foil bonded around the edge of the disc but not at the land — is easily read as insufficient power, which is why the reflex is to retune the sealer. Retuning raises energy and can appear to work at reduced speed while making the upper bound of the window unreachable at full speed.
Changing supplier without changing diameter changes nothing. A substantial share of "the liner does not work" conclusions are conclusions about a dimension rather than about a product. Two suppliers cutting to the same nominal diameter will produce the same overhang.
The diagnostic order we now use for any speed-dependent seal failure
Measure the sealing land on production parts, at three points, before discussing anything else. Confirm the liner diameter against that measurement. Only then look at power, speed, gap and torque. In this project the first step would have resolved the question in under an hour, and it is the step that neither previous investigation took.