Bottle Filling Line Compatibility: Match Bottles to Lines

A glass bottle is compatible with a filling line when its real, measured spread of height, body diameter, ovality, verticality, neck concentricity and finish dimensions fits inside four windows set by the machine: the star wheel pocket, the guide rail corridor, the filler's nozzle travel and the capping head's float. Compatibility is therefore a relationship between one bottle and one machine configuration at one speed, not a property of the bottle. The same container can run cleanly at eight thousand bottles per hour and fail at twelve thousand, because a faster line leaves each station less time to correct position.

What follows deals only with the machine interface. How tolerances are defined, written on a drawing and inspected across an order is handled under glass bottle tolerance standards for height, diameter and verticality, and the inside of the neck, meaning bore diameter, inner wall condition and plug or insert seating, is handled under bore finish. Both need to be settled first, since a machine window cannot be compared with a bottle whose dimensional system is still undefined.

Start from the line and derive the bottle windows

Most compatibility questions are asked in the wrong direction. Someone picks a bottle, asks whether it will run, and gets a reply based on a loosely similar container. A more dependable route is to take the windows off the machine first. They can all be measured, and none of them depends on which bottle you eventually choose.

Four windows cover nearly everything:

• Star wheel pocket. Pocket size and the clearance you intend to run give a permitted range of body diameter.
• Guide rails. The rail setting gives a body diameter corridor at the height where the rails touch the glass.
• Filler. Nozzle or valve travel gives a height window.
• Capping head. Float and chuck range give a window for neck concentricity and finish dimensions.

Put on one sheet, these are a bottle specification written in machine terms. Any drawing can be held against them at a desk before a sample is ordered, which sorts candidates into those that will very probably run and those that need a trial.

Do not treat the windows as permanent. Rails get moved, pockets wear and heads drift, so re-measure from time to time. A bottle rejected a year ago might pass now, though the opposite case, a bottle that used to run and no longer does, is the more frequent one.

Bottle properties each station reads

A filling line never assesses the bottle as a whole and never consults the drawing. It is a sequence of stations, each of which accepts the container only within a narrow band of position, size and shape. That explains why a bottle that looks faultless on the bench is thrown out five metres further along. What the machine actually touches is a contact band on the body, a height datum, a diameter across the pocket and the outside of the neck.

The unscrambler and infeed worm need a predictable body diameter and base so each bottle can be caught and spaced at a constant pitch. The star wheel pocket is a machined opening with fixed clearance, so it needs a consistent diameter in the band it grips. The rails define a sideways corridor. The filler works from the top of the finish, where the valve seals or the fill level is referenced. The capper works from the neck, which the chuck must centre on.

The bottle properties feeding those interfaces are:

• Overall height, taken from the base to a defined datum, not to the highest point of an uneven finish.
• Body diameter in the bands that meet the star wheel and rails, which may not be the widest part of the bottle.
• Ovality, or out of round: one diameter measured at different clock positions.
• Verticality: how far the body axis leans from plumb.
• Neck concentricity: the offset between neck axis and body axis.
• Finish dimensions, which govern capping.
• Brimful capacity and fill point, which govern declared net content.
• Bottom flatness, because a rocker base will not stand level on a conveyor.
• Mass distribution, since a heavy shoulder over a light base behaves differently on a fast transfer.

How each dimension goes wrong on the line

Height

Height looks harmless and is routinely underrated. Where fill is set by volume and the valve seals on the bottle mouth, the nozzle descends a fixed distance that is set once. Spread in height across an order alters how far the nozzle engages, which affects the seal at the valve and, on certain machines, the volume delivered.

The capper has the same sensitivity. A head that comes down to a fixed stop applies a different top load to a tall bottle than to a short one, so two containers off one pallet can leave with visibly different cap compression.

Body diameter and ovality

Diameter acts in two places. In the pocket, clearance controls how well the bottle is centred under the valve. With excess clearance the bottle drifts by a fraction of a millimetre, enough for a capping chuck to catch the thread off centre. With too little, the pocket grips hard: shoulders scuff, the transfer gets noisy, and thin walled containers break in a way that gets attributed to the glass when the pocket is at fault.

At the rails, corridor width decides between guiding and pinching. A gap set for nominal diameter pinches the largest bottles in the distribution and lets the smallest shingle, the term for bottles riding over one another and standing at an angle against the rail. Oversized bodies scuff; undersized ones shingle.

Ovality is not constant up the height of a moulded container. Many bottles are nearly round at mid height and clearly oval near the shoulder or base, where forming is still settling. An inspection that gauges only the middle will report a good figure while the shoulder contact band sits outside the window. Measure where the machine touches, not where the gauge fits easily.

Verticality

Lean is the property that speed punishes. A slightly leaning bottle stays upright on the conveyor, but each millimetre of lean is magnified over the bottle's height, so the finish swings sideways as the bottle turns. A slow capper tolerates this. On a fast line the head arrives before the bottle has settled, and you get a cocked cap or cross thread on perhaps one bottle in a thousand. That rate is enough to spoil a shift while never yielding a clear reject sample to investigate.

Neck concentricity and finish dimensions

Capping is where small bottle errors turn into visible defects on finished goods. A chuck or magnetic head has limited float, meaning the sideways mismatch it can absorb between its own centre and the neck. On most machines that float amounts to fractions of a millimetre. Once the neck offset exceeds it, the head cannot centre itself and drives the cap onto a thread that is not square to it. Typical results are a tilted cap, a closure thread cut on one side, a tamper band that splits at first opening and torque readings scattered well beyond the specification window.

The way the bottle is held makes this worse. Star wheel and rails both grip the body, and the capper normally depends on that same grip to keep the bottle still while the chuck turns. With an offset neck, the chuck aligns to one axis while the bottle is fixed on another. The difference is absorbed by whichever part is softest: liner, closure or the finish. A concentricity fault can therefore look like a sealing fault on a pack whose seal design is perfectly sound.

Finish height, thread outside diameter and thread root position together set how far the closure rotates and how hard the liner is compressed. They are dimensioned and inspected as a group for a reason. A finish at the bottom of its height tolerance paired with a closure at the top of its own can leave the liner scarcely touched, although each part passes individually. When capping goes wrong, measure the finishes of the bottles that failed, not of those that passed.

The transfer into the capper is a further, often missed, interface. If a rail pushes the bottle into position instead of a pocket indexing it, entry speed and rail pressure determine how much it rocks on arrival. A faintly domed base or small rocker makes the bottle arrive with a rotation the capper cannot detect. The defects come and go and are difficult to reproduce on a stationary rig.

Line parameters matched to bottle characteristics

The table is arranged by machine parameter because that is how the problem gets solved: begin with what the line is set to do, identify the bottle property it reacts to, then decide what to watch during a trial. Adjustments are given as a direction of change only, since ranges differ from machine to machine.

Measuring incoming bottles the way the machine does

When production bottles behave differently from the approved samples, the usual culprit is a mismatch in measurement, not in manufacture. Suppose the glass plant gauges height on a plate with a flat pad while your receiving inspector rests a caliper on the highest point of the finish. One bottle then yields two numbers. A sound batch can be quarantined and a doubtful one accepted, and the argument that follows has nothing to do with the glass. Use the datum the machine uses, with an instrument that contacts the bottle in the same manner.

• Height. Stand the bottle on a flat surface plate and lower the gauge onto a defined finish datum. A flat pad is preferable to a point, because a point reads a local irregularity instead of a dimension.
• Body diameter. Measure in the contact bands, at the height above the base where star wheel and rails touch, and take two readings at ninety degrees so ovality is recorded, not concealed.
• Verticality. Roll the bottle against a stop on the plate and read run out at the top with a dial indicator, or spin it on a mandrel and read sideways movement of the upper body.
• Concentricity. Rotate the bottle in a V block and read the neck outside diameter and the sealing surface.

Capacity needs two figures because they answer separate questions. Brimful capacity, filled to the very top of the finish, shows how much glass volume varies from bottle to bottle. Working capacity, filled to the intended point with the intended headspace, is what declared net content rests on. Because the fill point is measured from the base, variation in bottom thickness or push up can give a loose working capacity alongside a tight brimful distribution.

Log the conditions with the readings: gauge identification, calibration status, operator, temperature and the exact datum. Then give the readings a decision rule. A sampling plan with a stated sample size and acceptance number, referenced to a recognised scheme such as ISO 2859-1 and an agreed AQL, prevents every delivery from becoming a negotiation. Agreeing datum, instrument type and reading position in writing does more good than a tighter tolerance, because a tight tolerance measured inconsistently controls nothing.

Match the inspection plan to the packing format too. Bottles shipped as bulk-packed glass containers behave differently at the depalletiser from identical bottles on layer pads, and first contact damage is frequently done before the line has run one bottle.

Drawing limits versus the distribution the line receives

Two deliveries can both conform to the drawing and run nothing alike. A drawing sets limits; the line meets a distribution. One batch clustered at the tight end of the height window and another spread across all of it will both clear goods inwards, yet only the first runs without intervention. This is why a trial should be built around the extreme bottles instead of the typical one.

Shape matters alongside width. A batch skewed slightly tall loads a fixed stop capping head harder than its mean implies. A batch with two populations, which arises when two moulds or cavities behave differently, gives a line that runs perfectly, throws a sudden burst of defects, and repeats the cycle pallet by pallet. Requesting height, diameter and ovality as sets of readings, not as a pass or fail statement, exposes this before the bottles reach the infeed.

Keep what the line needs separate from what the drawing permits. If the line is comfortable over the whole drawing window, a tighter bottle buys nothing. If it runs only in the middle of the window, the first move is normally to change the line setting, not the bottle, at least until the cause is known. Tightening a specification to mask a drifted setting hides the real fault and makes the purchase carry it.

Changeover parts and bottle families

Provided the new bottle belongs to the same family as the old one, changeover is generally a much smaller job than buyers anticipate. Change parts divide into three groups, and identifying which group a proposed bottle change affects is the fastest way to size the work.

Sort the bottle range into families on paper before ordering anything. Two bottles with a common body diameter and finish but different heights normally need one set of adjustments and no parts. Where only the finish is shared, budget for pocket and rail parts. Where nothing is shared, plan a new line project, not a changeover. Grouping this way keeps the changeover kit small and the setting sheet short enough for operators to use.

Changeover is also a quality risk. Each one returns the line to a setting that was right months earlier and may since have been overtaken by wear: a nudged rail, a drifted head, a pocket insert replaced with a slightly different part. A brief first article check after every changeover, covering fill weight, cap torque and cap tilt on the first twenty bottles, catches most of this. For bottles that will bear load in a stack or beneath a capper, have compression behaviour on file before sign-off for a fast line; that is a separate procedure, described under the vertical load test for glass bottles.

What to assemble before a trial run

A trial settles one narrow point: whether this bottle, at this setting, keeps its position through a full loop at target speed. Three packages of information must be ready before the bottles arrive. Trials are most often wasted because only one of them was prepared.

1. Line data sheet. Speed in bottles per minute at the intended fill volume; fill method; number of filling valves and their pitch; star wheel pocket diameter and clearance; rail width and height settings; capping head type and float specification; cap feeder configuration; conveyor height and transfer geometry. A photo of every station with a bottle in it tells more than most drawings.
2. Bottle data. The toleranced drawing, finish designation, intended fill volume with headspace and, above all, the measurement records from the production run that will actually supply the bottles. A gold sample says nothing about a line, because lines are not fed gold samples. If the measured distribution cannot be supplied, request the span of the last twenty measurements instead of a single certificate value.
3. Sample set. Enough bottles for several complete loops plus spares for the breakage a trial causes, chosen to include the tallest and shortest, widest and narrowest, most and least oval. Ten bottles drawn from the centre of a batch will pass on almost any line, which will then go on to fail in production.

With line parameters and bottle specification side by side, we can judge compatibility at a desk: which properties the line is truly sensitive to, which the current bottle already meets, where the risk lies and what the changeover involves. The trial checklist then targets those specific risks, with named observations and speed steps, instead of testing everything at once. A marginal dimension is far better found on paper than at the capper at three in the morning.

Reading failures back to their cause

A failure shows at one station, but its origin is usually upstream or in the bottle. The patterns below lead to different corrective actions, and mixing them up means working on the wrong part of the machine.

Breakage is blamed on the glass more readily than any other fault and deserves the most scepticism. Glass breaking at one identical spot on the line every time is normally being forced by a setting. Breaks at random locations, or a check or splinter at the finish where nothing mechanical touched it, suggest a container condition. To tell them apart, log the exact position of every break for a whole shift. Clustered positions are a machine finding; scattered positions are a bottle finding.

Cosmetic damage is a category of its own because it never produces a broken bottle. A scuff, a light contact mark or a small shoulder chip can be made by the line, pass every functional test and still be refused by the customer on sight. Such damage accumulates and only shows after a long run, so it tends to be discovered in the warehouse. Look for it deliberately during the trial instead of waiting for a complaint.

Questions buyers ask about filling line compatibility

Will any glass bottle run on any filling line?

No, though the barrier is usually a setting, not something fundamental. Almost any bottle can be presented to almost any machine once pocket size, rail gap, nozzle travel and capping head height are set for it. The true exclusions are bottles whose diameter, height or neck geometry lies beyond the line's adjustable range.

Which bottle property gives the most trouble?

Verticality and neck concentricity, as a pair, account for more unexplained defects than anything else. Diameter and height faults are obvious and repeatable, so they are corrected quickly. A slight lean or neck offset shows up as tilted caps, torque scatter and occasional jams that come and go with the pallet, and is often misread as a closure or valve problem.

Does the fill method affect how much height variation is acceptable?

Substantially. With volumetric filling through a nozzle that seals on the mouth, height is usually the tightest dimensional requirement on the entire line, because stroke is fixed and the seal depends on where the finish really sits. A gravity filler with a level probe forgives height far more but becomes sensitive to working capacity. One bottle can thus be acceptable on one machine and marginal on the next.

Can a single changeover kit serve two bottle sizes?

Often, when both sit in one family. Group the range by diameter, height band and finish before ordering change parts; bottles sharing diameter and finish need only height adjustments, which is the least costly way to hold changeover time and spend down.

Why do bottles that pass incoming inspection still jam?

Goods inwards generally checks dimensions against limits, while the line meets a distribution and a contact geometry. A conforming batch may hold a skewed sub population from one mould, or meet the height limits with an uneven finish top that no single height reading reveals. The inspection datum may also differ from the machine's. Recording distributions instead of pass or fail closes most of the gap.