Forming Line Defects That Are Actually Ware-Handling Problems

Most plants I audit are convinced they have a mould problem. Half the time, they don't. They have a takeout, a dead plate and a lehr loader that nobody has recalibrated since the last cold repair, and a cold-end reject sheet that has coded the resulting damage as "forming" for years because nobody ever traced it back to where it actually happened.

That gap costs plants real money and it costs them the right fix. If your reject taxonomy can't tell a heel tap from a blank-mould check, you'll keep polishing plunger tips for a fault that lives 300mm downstream of the blank side.

Your reject sheet is lying to you

Industry process audits that break out reject codes by true root cause, rather than accepting cold-end operator coding at face value, commonly reclassify 25-40% of containers logged as "forming" or "mould" defects as ware-handling in origin. That's not a rounding error. On a mature line running above a 90% pack-to-melt target, a third of your forming rejects being misfiled means your entire improvement plan is pointed at the wrong equipment.

In 2014 I was running forming on a five-line plant in Brisbane when we chased a check problem for six weeks with nothing to show for it. The blank moulds looked clean, plunger cooling was in spec, cavity temperatures were tight. We finally marked ware coming straight off the takeout with a grease pencil and walked it down the line by hand. The checks were forming at the dead plate transfer, not the blank cavity. Not a mould problem. A transfer problem. Checks generated at the cross-conveyor or dead-plate transfer point are visually near-identical to blank-mould checks but originate 100-300mm downstream of the blank side, and if you're only reviewing mould inspection logs you will never find them. You have to mark the ware.

Three ware-handling faults wearing a forming defect's uniform

Takeout tong timing is the first one I check now. Tolerance is typically held to within ±10-20ms of the machine cycle, and drift beyond that window produces heel taps and push-up cracks that get miscoded as blank-mould wear on the reject sheet every time. The tooling didn't wear. The tongs are picking up a beat early or late.

Cross-conveyor speed mismatch is the second. Run section speeds of 100-200 containers per minute depending on double- or triple-gob tooling, and a conveyor running even a few percent slow will logjam ware into sidewall contact. Inspectors log that as annealing-lehr stone or checks nine times out of ten, because it shows up downstream and looks like a lehr problem. It isn't.

The third is cold-end coating timing. Hot-end tin or titanium coating goes on in the 550-600°C window right after forming. Cold-end polyethylene wax goes on post-lehr, below roughly 150°C. Mistime or under-dose that second coat and you get scuffing and line-contact abrasion that reads like mould-line wear on the inspection bench (and I've had fitters swear black and blue the mould is the problem right up until we pull the coating log and prove otherwise).

The lehr breaks the chain of evidence

Annealing lehr residence time runs roughly 40-60 minutes on a controlled cool-down from around 580°C to under 50°C. A ware-handling jolt at lehr entry, a pusher bar out of alignment, a star wheel that never got reset after the last job change, can seed a stress riser that doesn't fail on the line at all. It fails hours later, in the warehouse or on the customer's own filling line, as spontaneous breakage with no obvious cause.

And by the time that container breaks, the event that caused it is long gone from anyone's memory and every log. Nobody codes "lehr loader" against a pallet failure three states away. That's exactly why the lehr has to be treated as a diagnostic window, not a fixed 45-minute constant you set once and stop thinking about.

The defect doesn't lie about where it showed up. It lies about where it started.

Why the usual audit doesn't catch this

An OEM-affiliated forming consultancy will default to mould, plunger and blank-side diagnostics, because that's the equipment they hold service contracts and spare tooling for. A generic Lean or Six Sigma boutique will run standard SPC and a Pareto of defect codes without a glass-specific taxonomy, which means they take the cold-end coding at face value and never re-trace it. Neither one marks ware at the takeout. Neither one asks whether a "mould check" actually started at the dead plate.

This is the argument for bringing in an independent, vendor-neutral container glass consultant rather than an equipment vendor's field service team. A vendor-neutral audit has no incentive to find a mould problem, or a conveyor problem, or anything in particular. It has an incentive to find the actual one. On a line running an older Emhart 8-section against a newer Heye timing screw retrofit, the failure signatures are different enough that treating "ware handling" as one bucket misses the fix entirely.

Fix the taxonomy, then fix the equipment

This matters most in the first hour after a job change. Well-optimised IS lines are pushing mould, blank and neck-ring swap toward the sub-15-minute mark, but the guide rails, timing screws and star wheels reset in the same changeover are the most common source of the defect spike that follows. Plants blame "the new mould settling in." It's usually a loader or conveyor recalibration that got skipped because nobody owned it on the changeover checklist. Our Job Change Tool exists specifically to close that gap, mapping ware-handling resets into the same 9-stage Job Change Lifecycle as the mould and recipe change instead of leaving them as an afterthought nobody signs off on. It's not exciting work. It's the work that actually moves the pack-to-melt number.

In the US, that gap is getting harder to ignore. O-I's Fit to Win programme has concentrated volume onto fewer, higher-utilisation furnaces since 2023, which means every remaining line is running closer to its ware-handling limits with less slack to absorb a bad calibration. The Glass Packaging Institute has flagged the same thing from a different angle: a shrinking bench of experienced hot-end and ware-handling technicians as older operators retire, which GPI members repeatedly cite as a driver of preventable defect rates that never should have made it past the takeout.

Start by pulling your last quarter of reject data and asking a blunt question: how many of those "mould" codes were ever actually traced past the blank side? If nobody can answer that with certainty, that's the audit worth running before you touch a single plunger tip. Our forming audit starts at the takeout, not the blank mould, and our defect glossary keeps the visual distinctions between a mould-side check and a ware-handling check side by side, so your cold-end crew stops coding one as the other.