IS Machine Section Timing: The Case for Locking SKU Profiles

It is 2am on a Tuesday and section 4 on your 10-section line has just cleared a jam. The machine setter is back at the keyboard, pulling the timing from memory, or from a sheet taped inside the cabinet door, or from a photograph on his phone of last month's printout (I have seen this on three separate audits in the past two years). The section comes back up. And for the next 40 minutes, your warm-end inspector watches bird swing and leaners roll through because nobody locked the reheat time when the job changed.

That's not a manning problem. That's a systems problem.

The re-discovery window is what you're not measuring

Most container glass plants in the GCC run IS timing calibration from scratch on every production run. In 2017 I was auditing a three-furnace plant in the region, strong crew, experienced setters, a well-maintained IS machine fleet running Emhart AIS controls. The average window from section restart to stable pack was 52 minutes per section. Fifty-two minutes. On a section the crew had run dozens of times that year.

Operator-reported changeover scrap windows of 45–90 minutes are typical in GCC plants where timing is re-calibrated per run, compared to 15–25 minutes in benchmarked European plants. That gap is a direct consequence of non-locked SKU timing profiles, and it is remarkably consistent across every plant I audit in the region.

Plants without locked profiles typically show section-to-section weight distribution variance of 4.5–7.0% of nominal in the first 30 minutes post-changeover, generating 200–400 reject containers per section before stabilisation. Run that across 10 sections twice a week and it stops being a forming statistic. It shows up in your cost-per-tonne.

The parameters that will not tolerate estimation

Not all IS machine timing parameters carry equal penalty for being wrong. Three matter most.

Reheat time, the free-air interval between blank invert and blow-mould close, has a permissible window of ±0.05 s around the SKU-specific optimum. Go +0.1 s long and you get parison sag, fold or wrinkle defects that track through QA into packs. Go -0.1 s short and the neck is still viscous when the blow mould closes, producing bird swing, the glass fin at the parting line that triggers a section split and a reject spike before the inspector logs the alarm.

On NNPB lines, plunger contact timing must land within ±0.03 s. Emhart AIS and Heye CIS controllers operate at 10ms timing resolution, and mis-timing by more than 20ms produces off-centre finish, the lean defect that sits in the top five reject modes on beer and carbonated-beverage bottles. This is not a parameter you estimate from how the last run felt. It is a number from the SKU profile.

Settle blow pressure and duration are equally SKU-specific. For Blow-and-Blow, the target is 0.07–0.12 bar for 0.35–0.55 s, narrowing or widening with neck diameter. A universal settle blow setting applied across a mixed SKU family inflates the warm-end check defect rate by +0.8 to +1.4 percentage points versus a locked profile. Most setters think they compensate by eye. The warm-end data rarely agrees.

The timing profile is not a starting point for the setter's judgment. It is the answer. The setter's job is to protect it, not re-derive it on every run.

What actually breaks a locked profile

Gob weight variance breaks it first. When section sigma exceeds 1.5 g per hour (target: 0.8 g or less), the counter-blow timing set for the nominal gob starts producing choke defects on light gobs and splits on heavy ones. Not a timing failure. A gob control problem that has migrated downstream into your timing parameters. You cannot lock timing reliably until gob weight is stable. This is the dependency most plants miss.

Cold section restart is the second break point. Blank mould temperature should sit at 430–480°C at end of blank cycle, measured by contact pyrometer or a Williamson non-contact unit. When a section comes back cold after a jam, apply the locked profile immediately and you have re-entered re-discovery. You need a 15–20 minute warm-up buffer first. The profile is still right. The mould is not ready for it yet.

But the break point that surprises most people is personnel. The machine setter holds the timing parameter file and is the gatekeeper of the SKU profile. In plants without a formalised library, that knowledge lives with the person, not the system. One shift change or one sick day at startup and the incoming setter is re-discovering timing that was already found and validated on the previous run. Look, most setters are good at their jobs. The problem is a system that forces them to carry knowledge in their heads instead of a file. In GCC plants I have audited, that scenario extends re-discovery by 20–45 minutes and adds 0.3–0.7 percentage points to that run's reject rate.

Lock it once, version it, carry it with the mould set

Every IS section timing parameter, settle blow duration, reheat time, counter-blow timing, plunger contact timing, takeout advance, should be captured at the validated optimum for each SKU, version-controlled, and locked to the mould set as part of the forming recipe. When the mould set moves, the timing profile moves with it.

Verallia made this case publicly in its 2023 Universal Registration Document: a 2.1% warm-end cullet rate across 33 European furnaces, with IS machine timing standardisation named as a lever in its 'Objectives 2025' programme targeting a 0.3 percentage-point improvement. A company of that scale, with that depth of process culture, still had timing lock as an open efficiency item. That should tell you how widespread the gap is.

The SKU Library component of Lean Glass's Job Change Tool operates on this architecture: recipes, mould sets, and forming specifications version-controlled and locked, with change-control sign-off before a setter can deviate from the profile. The vendor-neutral design matters here because this fix is not tied to any controller platform. Emhart, Heye, BDF. The discipline is identical across all three.

If you want a structured approach to IS machine timing optimisation before committing to a full change programme, a forming audit covers section-level timing variance, mould temperature profiles, and the gap between what setters are running and what the best-known-state profile specifies. For plant managers and container glass consultants working with a GCC or European fleet, it is the natural starting point.