Glass Bottle Capping Machine Setup and Torque

Setting up a capping machine for a glass bottle starts with the closure, not the torque dial: the cap family fixes which head and motion you need, and the finish diameter, for example 28 mm or 38 mm, fixes the tooling. Torque comes last. You find it by running empty bottles slowly, then empty bottles at production speed, then filled bottles across a deliberate spread of settings, and you confirm the result with an integrity check on the filled pack. Any value you arrive at is a reference range for your own head, closure and line speed; it does not carry over to another installation.

Four closure facts that select the head

The head is indifferent to the bottle's capacity, colour or body shape. What it responds to is the closure and the way that closure holds on to the glass. Four pieces of information settle the choice.

• Closure family. Crown, continuous thread screw cap, push-in plug and child-resistant cap are each retained by a different mechanism. Moving between families normally means fitting another head to the same machine frame, not re-adjusting the one that is there.
• Diameter. This is given as the neck finish designation, such as a 28 mm or 38 mm finish. It determines the gripping or crimping diameter, so going down a size requires change parts.
• Engagement geometry. For a threaded cap, pitch, thread form and the count of thread starts govern how far the cap drops per turn, and so how many turns and how much vertical travel the head has to provide. For a crown, skirt diameter and the bead on the glass lip govern crimp depth. For a plug, the interference between plug diameter and finish bore governs the insertion stroke.
• Sealing element. A cap liner, a plug gasket and a vacuum seal each ask something different of the closing motion. A setting that seals well on compressible foam may crush a stiff pulp liner or leave a plug gasket short of its seat, and the machine cannot tell which it is working on.

With all four known, the head family is as good as decided. With one missing, the answer gets worked out by trial on the line, which is the costliest place to learn that head and closure disagree.

Head, motion and settings for each cap type

Each closure family needs its own force profile, and most adjustment trouble makes sense once that profile is clear.

A crown is crimped. The head comes down, seats the crown on the lip, and jaws or a rolling ring fold the skirt beneath the lip bead. The movement is mostly axial with a brief radial forming step at the end. Depth of crimp, and how evenly it repeats around the circumference, is what counts: a shallow crimp lets go under pressure, a deep one loads the lip and may chip or crack it. If the pack is a crown going straight to filling, read up on how the crown interface seals on glass before touching the crimping head.

A continuous thread screw cap is spun on by a rotating head while the bottle is clamped so it cannot turn. Because the thread draws the cap downward, rotation has to be paired with a controlled downward preload. Grip force, turns, applied torque and the stop condition are the settings. Stopping on a count of turns gives varying seating when thread start positions differ from bottle to bottle; stopping on torque gives consistent load but can be fooled when friction changes.

A push-in plug goes in on a straight stroke into an interference fit, with no rotation at all. Depth, speed and the alignment of plug to bore are the settings. A plug that enters off-axis sits cocked and leaks, and extra force does nothing to cure it.

A child-resistant cap, push-and-turn or squeeze-and-turn, closes like a screw cap, except that the head has to drive through a slipping or ratcheting element inside the cap. That element soaks up part of the closing torque. A reading taken through the cap during capping therefore describes the mechanism as much as the seal, so this family is checked by function as well as by number.

A metal lug lid closed under vacuum is a separate case. The equipment is a lug roller or seamer, and internal vacuum makes the seal, not a closing load. Thread-head reasoning does not apply, and the conversation belongs with the seamer supplier, not with whoever supplies screw heads.

Finding the torque setting in four stages

Two settings on a screw head are often mixed up. One is the stop condition: the rule by which the head judges the cap closed, whether a turn count, a torque, a vertical travel or a mechanical stop. It is fixed once the head is correctly tooled. The other is the torque target reached at that stop, which is what creates sealing compression in the pack.

The search moves from the least realistic pack to the most realistic, since every stage strips out one variable and costs less than the stage after it.

1. Empty bottles, low speed. The aim is not a final torque. You are confirming that the cap starts properly, the thread picks up without crossing, the head grips and lets go cleanly, and the cap lands square on the finish. Torque is read only to see that something is being applied and that it repeats from bottle to bottle. Cross-threading here points to alignment, thread start position or tooling, and no torque value will remove it.
2. Empty bottles, approaching production speed. This exposes speed effects. A head that seats well when slow may release the cap before the thread has fully engaged, or the bottle may not be clamped firmly enough. You see caps that are square but loose, or tight in one part of the batch and loose in another. Solve these mechanically before tuning torque; using torque to cover a speed fault yields a batch that passes at the plant and leaks in distribution.
3. Filled bottles, production speed. The liner now works against real product. Apply torque over a deliberate spread, from plainly too low to plainly too high, and read removal torque on each group after a fixed conditioning interval. What comes out is a window in which the pack seals yet still opens comfortably.
4. Integrity check on the filled pack. Torque only controls the compressive load indirectly. A damaged liner, a chipped finish or a cocked cap can leak at a perfectly good reading, so the integrity check is what proves the setting does what the number suggests. It closes the sequence; it does not replace it.

How application and removal torque differ, how each is measured, and how a window is written into an incoming inspection standard are set out in our page on torque testing and acceptance for bottle caps. Use that to set the target and this sequence to make the machine deliver it.

How finish and closure tolerances shift the settings

Two packs with the same specification sheet can call for different head settings. The cause is the tolerance stack between glass and closure: the machine must cope with whatever spread of finishes actually arrives, and that spread is never one value.

On the glass side, thread outside diameter, thread start position, sealing band or bore diameter, and finish height all move within an agreed band. A head set to the nominal figure, and not to the centre of what was delivered, passes some bottles and fails others. The terms behind those dimensions are explained in our guide to how a glass thread finish is defined and drawn.

On the closure side, moulded caps differ in skirt diameter and internal thread form, and lined caps differ in liner thickness and seating depth. A liner at the thin end needs more turns, or a little more torque, to reach the compression a thick one reaches sooner. Change the liner grade or the cap mould and yesterday's setting may sit outside the new window, though bottle and cap designation are the same on paper.

Together the two stacks define how much adjustment the head must offer. A wide combined stack demands enough travel and torque authority to close the loosest pairing without over-closing the tightest. A narrow one allows a lighter head and a smaller window, and the pack runs more forgivingly. It pays to quantify the stack before equipment is chosen.

Three checks make this routine:

1. Measure the finish on a representative sample of the delivered batch and compare the spread with the drawing tolerance.
2. Close that sample with the production closure and look for squareness and skirt contact right round the rim.
3. Note the head setting at which every bottle in the sample passes and compare it with the previous batch. A shift between batches tells you about the incoming glass or the closure lot, not the operator.

Defects and the order to correct them

The sequence of correction matters more than any single fix. Adjust out of order and one fault ends up hidden behind another, which is how a line acquires a torque window wide enough to conceal a real leak.

Cocked or tilted cap

This is an alignment fault. Usual sources are a bottle not held square, a head off-centre to the bottle axis, the wrong support height, or a thread start out of step with the descending head. More torque crushes the liner on one side, opens a gap opposite and may damage the finish. Correct bottle position and support height, then concentricity, then thread start timing, and leave torque until last.

Cross-threading or a stripped thread

This is an engagement fault. The cap is being turned before the thread starts have met, the head is descending quicker than the thread can draw the cap down, or rotation begins before the cap is square. Timing and head speed are the remedy. Extra torque just shears the thread sooner.

Leak at the correct torque

The load was applied but no seal formed, which makes this the most informative defect. Suspect a damaged or displaced liner, a chip or poor sealing band on the finish, a cap that is square yet not fully down, or a setting that suits a different liner grade. Check liner, finish, seating and torque window in that order.

Cap too tight to open

This is over-compression. Lower the application torque or change to a liner that seals at less load. Tightening to cure a leak is the wrong direction once the seal threshold is known: excess load cold-flows the liner off the rim, can split a plastic skirt and can chip a thin finish, so the pack may lose its seal just as it becomes hard to open.

Torque present on some bottles and missing on others

Where there is no pattern, look at grip force, a worn clutch in the head, or uneven bottle clamping. These are wear and maintenance matters. Read a run of consecutive bottles and look for a drifting trend as opposed to random scatter.

Changeover checklist

Format changes account for most avoidable downtime. Do the mechanical work before the numerical work, in this order.

1. Identify the closure family of the new format and the head family it needs. If the family is different, change the head and not merely the tooling.
2. Fit tooling for the closure diameter and confirm that gripping or crimping parts are inside their wear limits; worn parts close inconsistently however good the parameters.
3. Set the cap chute, escapement and any orienting or sorting parts, and check each cap arrives square and the right way up. A tilted cap cross-threads under any head.
4. Set support height and clamping for the new bottle so it resists rotation without being distorted.
5. Check head concentricity to the bottle axis, then thread start or crimp alignment timing.
6. Enter the stop condition and torque target from the measured window for this particular bottle, closure and liner, treated as a reference range to confirm on the line. Never carry a figure over from the last format.
7. Run empty at low speed, empty at production speed, then filled at production speed, reading torque and seating each time.
8. On the first samples, measure removal torque after the fixed conditioning interval and inspect seating round the whole rim before the batch is released.
9. Log the passing settings with the closure lot and glass batch so the next changeover begins from a known point.

One piece of work comes before all of this. Whether the cap physically seats on the neck, and what a click or a wobble means, is established by a cap fit test recorded as pass or fail. A head cannot be tuned into a pack whose cap does not fit the bottle.

Monitoring torque drift during a run

A setting that passes at changeover will not hold itself for a shift. Heads drift, clutches wear, belts slacken, caps vary by lot and bottles vary by pallet.

• Take removal torque on a defined sample at a defined interval, always with the same conditioning interval and grip method, so readings can be compared.
• Plot the values. A trend toward one edge of the window is the early warning, and it disappears if only pass or fail is written down.
• Inspect seating alongside each reading. Square, fully seated and in-window is a healthy pack. A square cap at the window edge, or a normal reading on a cap that is visibly not down, both mean the head needs attention though nothing has formally failed.
• Keep settings logged per format and compare the log with the torque trend. If the setting needed to hold the same torque creeps, the cause is head wear or a new closure or liner lot, and knowing which decides between maintenance and a call to the closure supplier.
• Hold back samples from each batch under the intended storage conditions and read them again after the distribution cycle. Removal torque falls with transport vibration and liner relaxation, so the plant reading is the optimistic one and the later reading is the honest one.

Questions about capping machine setup

Can one machine apply both crowns and screw caps?

Seldom with a single head. A crown is formed by an axial crimp and a screw cap is spun on with preload and a stop condition. Some frames take interchangeable heads, so the frame may be common while head, tooling and parameter set are not. Confirm which head families a machine supports before counting on it for both formats.

Why does removal torque fall in transit?

Stress relaxation in the liner lowers the axial load over time, and vibration lets the cap creep round in the opening direction. Retained samples re-read after the distribution cycle show whether the window had enough margin.

How often should torque be checked in production?

At a set interval and sample size, with an unchanged method, and recorded as a trend. Add a seating inspection each time, because a normal number on an unseated cap is a fault the reading cannot reveal.

What information is needed before setup can begin?

Closure family and diameter, finish designation, liner or sealing element, line speed and bottle support dimensions, plus the head family already installed if that is fixed. From these follow the head family, the tooling, a sensible stop condition and the span the torque search must cover. When we review a bottle and closure pairing for a project, these are the items we ask for first; if cap-to-neck fit is still unconfirmed, the fit test comes before any machine work.