PET Preform Selection: Weight, Neck Finish & Bottle Design

A PET preform is specified by three numbers that have to agree with each other: its gram weight, its neck finish (the moulded thread and support ring your closure mates to), and the bottle geometry it has to blow into. Get the neck finish wrong and your caps will not seal — a PCO 1810 cap on a PCO 1881 neck fails outright because the threads are cut to different pitches. Get the weight wrong and you either waste resin or blow a bottle that buckles. This guide sets out how to read and specify each parameter so the preform, the cap, the blow mould and the filling line are all talking to one another.

The short answer for most beverage buyers: confirm the neck finish standard by name and number (e.g. PCO 1881, 28 mm; or GME 30.21 for water), then fix the preform weight from your target bottle weight and stretch ratio, and only then discuss bottle shape. Below we explain why, with the dimensions that matter.

What a preform actually is

Injection-moulded PET preforms are the test-tube-shaped parts that a stretch-blow-moulding (SBM) machine reheats and inflates into a finished bottle. The neck — threads, support ring and sealing surface — is moulded to final dimensions at the preform stage and is not stretched during blowing; it is the only part of the preform that ends up at the same size on the bottle. Everything below the support ring (the body and base) is biaxially stretched. That single fact drives the whole selection logic: the neck is a fixed, standardised interface, while the body is a design variable governed by stretch ratio.

The three jobs a preform has to do

• Carry a standard closure interface. The neck finish must match the cap, the capper’s chuck, and (for aseptic or hot-fill lines) the sealing method.
• Deliver enough material, distributed correctly. Wall thickness in the preform body becomes wall thickness in the bottle after stretching.
• Survive the process and the product. Intrinsic viscosity (IV), drying and crystallinity at the neck determine whether the part blows cleanly and holds pressure or hot liquid.

Neck finish: the interface you must name, not guess

The neck finish is a published standard. You specify it by its name and number, and a compliant preform is one whose neck dimensions fall inside that standard’s drawing tolerances. The two families a beverage buyer meets most often are the PCO necks (PCO = “Plastic Closure Only”) used for carbonated soft drinks (CSD) and the lighter GME water necks used for still water.

PCO 1810 vs PCO 1881 — the comparison that trips up buyers

PCO 1810 is the older 28 mm CSD standard. PCO 1881 is the lightweighted 28 mm successor, standardised through the beverage industry to cut neck weight while holding pressure performance. The differences are specific and they matter on the line.

Sources for the dimensions above: Frystal Pet — PCO 1810 vs 1881 and the industry overviews at PAGpackaging and bottlepreform.com.

Two practical consequences follow:

1. Caps and necks are not mixable. A PCO 1810 cap on a PCO 1881 neck (or the reverse) gives a catastrophic sealing failure — the finer 2.70 mm pitch on the 1881 simply will not engage an 1810 cap cut for 3.18 mm. When you switch standards you switch the cap supply, the capping torque settings and often the capper change parts at the same time.
2. The lightweighting is real but bounded. PCO 1881 removes roughly 1.3 g of resin from every neck versus 1810. Across millions of units that is a material saving, which is why most new CSD lines specify 1881. But the saving lives in the neck; it does not by itself reduce body weight, which is set separately (see below).

When you should still choose PCO 1810: your filling and capping line was built for it, your existing cap inventory is 1810, and you are not planning a change part investment. Choosing for compatibility with installed equipment is a legitimate engineering decision, not a step backwards.

When you should choose PCO 1881: a new line, a high-speed line, or any programme where neck-weight saving and the shorter, faster 1.5-turn seal matter.

The history behind the 1881 switch

PCO 1810 was the long-running CSD standard for decades — a 28 mm finish with a generous neck that gave reliable sealing on a wide range of cappers. As resin costs and sustainability pressure rose through the 2000s, the beverage industry moved to a shorter, lighter neck that kept the same 28 mm nominal size and pressure capability while shaving roughly a quarter of the neck mass. The “1881” and “1810” are simply the reference numbers for those two published finishes; both are 28 mm “PCO” (Plastic Closure Only) standards, but their thread profiles, heights and matching closures are different parts. The reason the migration matters to a buyer is supply continuity: caps, slitting/folding cap tooling, and capper change parts are all standard-specific, so the decision ripples through your whole closure supply chain, not just the preform.

What stays the same and what changes

This table is the practical takeaway: a neck-standard decision is a closure-supply decision first and a resin-saving decision second.

Water necks and other finishes

For still water, the 28 mm PCO neck is usually overkill — water carries no internal CO₂ pressure, so the heavier pressure-rated neck wastes resin. The industry uses lighter standardised water necks instead, most commonly the GME 30.21 / 30.25 / 29-21 (PCO 1810 short) family and 26/22 GME necks. The 26/22 designation comes from the German mechanical-engineering association (Verband Deutscher Maschinen- und Anlagenbau), where the numbers refer to outer (~26 mm) and inner (~22 mm) neck diameters; it is a very light water finish. Larger 38 mm three-start necks are used for juice, dairy and isotonics where a wide mouth aids filling and pouring. These finish data sheets are published by CETIE (the European bottling-technology body) and the ISBT in the United States.

References: ISBT ThreadSpecs (voluntary beverage finish guidelines) and CETIE finish data sheets (GME). Always pull the actual dimensioned drawing for the exact finish before tooling a cap or buying a preform.

How to read a neck-finish drawing

When a supplier sends a finish drawing, five dimensions decide whether your cap and capper will work with it:

• T — thread outside diameter. The headline number; sets the cap’s internal thread diameter.
• E — thread root / inside-of-thread diameter. Controls thread engagement depth.
• I — bore (inner) diameter. The mouth opening; affects fill nozzle entry and pour.
• H — finish height. From the top sealing surface to the underside of the support ring; this is where 1810 and 1881 differ most.
• Support ring (flange) diameter and position. The capper grips and the line conveys the preform/bottle by this ring; air-conveyor and neck-handling rails are built to it.

A “compliant” preform is one whose measured T, E, I, H and support ring fall inside the published tolerance band for that named finish. Two preforms can both be “28 mm” and still be incompatible if one is 1810 and the other 1881 — which is why you never specify a finish by nominal size alone. Always specify the standard name and number, and verify incoming parts against the drawing with a neck gauge or by checking the supplier’s dimensional report.

Why the wrong finish causes rejected shipments

In practice the costly failures are not exotic. They are: a cap that strips or leaks because the pitch did not match; a support ring out of tolerance so the air conveyor drops bottles; a sealing surface with a moulding defect so aseptic seal integrity fails; or a neck bore that the filler nozzle fouls. Every one of these is a finish problem caught too late. The cheapest place to catch it is on the drawing and the first-article sample, not on the line.

Preform weight: derive it, don’t pick it

Preform weight is the lever that controls bottle wall thickness, top-load strength and resin cost. It is not chosen by feel; it is derived from the bottle you have to make.

The stretch-ratio rule

In stretch-blow moulding, the preform body wall thickness becomes the bottle wall thickness after biaxial stretching. The relationship is straightforward:

preform wall ≈ bottle wall × biaxial stretch ratio

So a target bottle wall of 0.5 mm at a biaxial stretch ratio of 8 needs a preform body wall of roughly 4 mm. This is the core sizing identity used in preform design (All Right Machinery — Preform Design).

Stretch ratio itself is the product of the axial stretch (how much the preform lengthens) and the hoop/radial stretch (how much it expands in diameter):

• Overall (biaxial) stretch ratio: commonly in the 18:1 to 25:1 range for 250–850 ml containers, and around 18:1 to 20:1 for 250–1000 ml (SciRP — Stretch-Blow Molding parameter prediction).
• Axial stretch ratio: preferably above ~1.7, more typically in the 2.2–3.2 range.

A correctly oriented PET wall — where stretching pushes the material out to meet the mould surface — gives the bottle its strength, gas-barrier and clarity. Under-stretch leaves thick, hazy, weak walls; over-stretch thins the wall and risks pearlescence or blow-out. Final container average wall thickness for beverage bottles typically lands in the 0.25–0.8 mm band.

Worked example

The exact gram weight is then confirmed by the preform supplier’s mould, because length, taper and base geometry also add mass. The point is that the number is engineered backwards from the bottle, not guessed. Lightweighting a bottle programme means re-running this calculation — usually trimming body weight while keeping enough wall to pass top-load and (for water) drop tests.

Where the weight goes wrong

Two failure modes bracket the weight decision:

• Too light (over-stretched). Pearlescence (a milky, stress-whitened look) in highly stretched zones, thin spots in the base, poor top-load so bottles telescope on the pallet, and — for CSD — creep and stress-cracking under pressure. The base and the shoulder are usually the first areas to fail because they see the highest stretch.
• Too heavy (under-stretched). Thick, hazy walls, poor gas barrier (CSD loses carbonation faster, sensitive products oxidise), wasted resin and slower cooling cycles. Under-stretched PET has not built the molecular orientation that gives the wall its strength, so paradoxically a heavier bottle can perform worse on barrier and burst.

The target is “fully oriented” PET — stretched enough that the material strain-hardens and meets the mould wall everywhere. That is why weight cannot be cut in isolation from the bottle shape and the SBM process window.

A note on rPET

Recycled PET (rPET) shifts the weight and IV decision because reclaimed flake usually has a lower and more variable IV than virgin resin, and carries more acetaldehyde and colour. Programmes running high rPET fractions often lift the preform weight slightly for process robustness, tighten the IV spec, and add an acetaldehyde-scavenger masterbatch to protect taste — see Masterbatch and colourants for food packaging. The neck finish itself is unaffected by rPET content.

IV: the material spec that sits under the weight

Preform weight tells you how much resin; intrinsic viscosity (IV) tells you which resin. IV measures PET molecular weight and governs melt strength, blow performance and burst/creep resistance. Water and CSD preforms typically use 0.76–0.84 dl/g bottle-grade PET, with higher IV for pressurised CSD and hot-fill. Specifying weight without IV is half a spec. We cover the IV bands in detail in our companion guide, PET resin grades by IV.

Bottle design: where neck, weight and shape meet

Three design features decide whether a derived weight actually performs:

• Base design. CSD bottles use a petaloid (footed) base to resist internal pressure; still-water and hot-fill bottles use champagne or heel-vented bases. The base is the hardest area to blow evenly and often the thinnest — base weight and stretch must be checked, not assumed.
• Panels and ribs. Vacuum panels absorb the volume change in hot-filled or pasteurised products; pressure ribs and a sound shoulder carry top-load in stacked pallets. These features change how material distributes during blow and can force a small weight increase.
• Label and grip zones. Shrink-sleeve and wrap-around labels constrain shoulder and heel geometry; deep grips concentrate stretch and need local wall control.

Hot-fill, aseptic and CSD change the rules

The neck region behaves differently by process:

• CSD: the neck must hold carbonation pressure; PCO finishes are pressure-rated for this reason.
• Hot-fill (~85 °C+): the neck is often crystallised (whitened) to resist deformation at fill temperature, and the bottle carries vacuum panels.
• Aseptic: the sealing surface tolerance is critical because the cap seal is the sterility boundary; finish dimensions and sealing-surface flatness are checked closely.

Match the finish and base to the process before you optimise weight.

A selection sequence you can follow

Putting the three parameters in order avoids re-work:

1. Fix the process and product. Still water, CSD, juice, hot-fill or aseptic. This decides pressure rating, whether the neck is crystallised, and the IV band.
2. Choose the neck finish by name and number. PCO 1881 or 1810 for CSD; a light GME water finish for still water; 38 mm 3-start for juice/dairy. Confirm the cap and capper match.
3. Set the target bottle: volume, body wall, base type, panels, label.
4. Derive the preform body wall from bottle wall × stretch ratio, keeping the overall ratio in the 18:1–25:1 window.
5. Confirm gram weight and IV with the supplier against the full preform geometry, then validate with first-article blow trials (wall-thickness map, top-load, burst/drop, and — for aseptic — seal integrity).

Skipping straight to “what does a 20 g preform cost” without steps 1–4 is how buyers end up with parts that quote cheaply and fail on the line.

How Innovote sources this

When a buyer comes to us with “I need preforms,” we turn that into a buildable spec before we quote, because an under-specified preform is the single most common cause of a rejected shipment or a line that will not run. Our intake checklist:

1. Neck finish, by name and number. PCO 1881, PCO 1810, a named GME water finish, 38 mm 3-start — with the dimensioned drawing or a sample cap to verify against. We confirm cap compatibility before tooling anything.
2. Preform weight and IV. Either you give us the gram weight and IV, or you give us the bottle (volume, target weight, fill process) and we back-calculate the preform body wall from the stretch ratio and propose a weight and IV band.
3. Bottle drawing or sample. Base type (petaloid vs heel-vented), panels, shoulder, label window — so the body weight and base weight are checked against the real geometry.
4. Process and product. CSD vs still water vs hot-fill vs aseptic — this fixes whether the neck needs crystallising and whether you need a higher IV.
5. Volumes, MOQ and packaging. Cavity count, mould compatibility, bulk vs bagged, and the landed-cost path into Egypt.

We document the finish standard and IV on the spec sheet and request the supplier’s technical data sheet and certificates of conformity to the relevant food-contact requirements — phrased as compliant with / meets the requirements of, certificates and specs available on request, never as a blanket “approval.” For the food-contact side of resin selection, see Food-grade vs food-safe resins. Tell us the spec and we come back with grade, IV, neck finish, MOQ, lead time and a landed-cost path.

FAQ

Can I use a PCO 1810 cap on a PCO 1881 bottle?
No. The thread pitch differs — 3.18 mm on 1810 versus 2.70 mm on 1881 — so the cap will not seal. Switching neck standards means switching the matching closure and capper settings together.

How much lighter is a PCO 1881 neck than a PCO 1810 neck?
The neck weight drops from roughly 5.1 g to about 3.8 g, a reduction of around 20–30%. The saving is in the neck only; body weight is set separately by the bottle’s stretch ratio.

How do I work out preform weight from my bottle?
Start from the target bottle wall thickness and multiply by the biaxial stretch ratio to get the preform body wall (e.g. 0.5 mm bottle wall × stretch ratio 8 ≈ 4 mm preform wall). The supplier then confirms the gram weight from the full preform geometry. Typical beverage stretch ratios run 18:1–25:1 overall.

What IV should a beverage preform be?
Bottle-grade PET for water and CSD is commonly in the 0.76–0.84 dl/g range, higher for pressurised CSD and hot-fill. IV is a separate spec from weight and both should appear on the data sheet. See our PET resin IV guide.

Which neck finish should I use for still water?
A lightweight water finish such as a GME 30.21 / 29-21 or 26/22 GME, not a pressure-rated PCO neck — still water has no internal pressure, so a CSD neck wastes resin. Always confirm against the published CETIE/ISBT drawing.

Do hot-fill bottles need a special neck?
Often yes — the neck is crystallised (heat-set, visibly whitened) so it does not deform at fill temperature, and the bottle body carries vacuum panels. Confirm the finish and process together.

Need preforms specified and sourced? Tell us the neck finish, target bottle and fill process — we’ll come back with grade, IV, MOQ, lead time and a landed-cost path into Egypt.

Related: Food-Grade Packaging Resins (hub) · PET resin grades by IV · Masterbatch and colourants for food packaging

By the Innovote Trade Desk.