Polyethylene terephthalate (PET) film enters the converting plant as a semifinished material whose primary specification is not chemistry alone but dimensional and surface-state geometry. The converter measures roll width to match slitter knives and press infeed systems, inspects core internal diameter to control unwind brake torque and chuck engagement, and identifies the corona-treated side to position the functional surface against printing, coating, metallizing, or laminating rollers. Because biaxially oriented PET film is produced by sequential or simultaneous stretching followed by heat setting near 200–230 °C, the base film surface is relatively low in wetting tension unless a corona discharge or chemical primer is applied. When a treated polyethylene terephthalate film roll arrives at a converting line without side orientation clearly documented, the downstream defect is rarely detected until adhesive or ink has already been applied. These three variables—roll width, core geometry, and treatment side—operate as a coupled system: a narrow slit roll wound on an undersized core will telescope under braking, and a correctly treated outer surface will be contaminated if the roll is unwound with the backside contacting idler rollers.

PET Film as a Converter Sees It: Roll Width, Core and Treatment Side

For converters, the phrase “roll width, core and treatment side” represents the incoming inspection point at which most downstream waste is predetermined. Width affects side-lay registration on narrow-web flexographic presses and gravure print repeats. Core internal diameter—commonly 76.2 mm or 152.4 mm—determines whether an unwind shaft can be expanded without core crushing. Treatment side orientation, specified as “inner treated” or “outer treated,” sets which surface receives the first coating. A master roll of PET film with a width of 1,220 mm and a core ID of 152.4 mm may be slit into 40 mm rolls on 76.2 mm cores. The purchase order must state the treated side; otherwise the converter cannot know whether the corona discharge was applied to the air side or the quench-drum side after orientation. This is not a cosmetic distinction. For metallization, a treated outside surface will accept vapour-deposited aluminium, while an untreated reverse side will exhibit poor metal adhesion and high optical density scatter. The same logic applies to solventless laminating, where the adhesive must be applied to the corona-treated side to achieve a laminate bond strength above 3 N/15 mm when tested according to ISO 11339:2022.

In slitting, the master roll width is not a free variable. The converter selects a slit pattern that minimizes edge trim and matches the repeat length of downstream presses. For a 1,600 mm master roll running on a duplex slitter at 250 m/min, the cumulative width of all slit rolls plus 2 mm edge trim must not exceed the measured master width by more than 0.5 mm. Width deviations on individual slit rolls are commonly specified at ±0.5 mm for rolls below 500 mm and ±1.0 mm for wider rolls, although the exact value depends on the substrate thickness and the cutting method. Razor slitting of film below 50 µm produces a low-dust edge but is sensitive to blade angle; rotary shear slitting of thicker PET film above 125 µm reduces edge cracking but requires higher rewind tension. The converter measures actual roll width with a calibrated steel rule or laser micrometer and records the average of three readings per roll. If the side-lay variation exceeds 1.0 mm on a narrow-web press, the registration system may not correct at speeds above 200 m/min. Width defects are rarely recoverable: once a roll is slit under-width, the entire set must be downgraded or sent to a rewinding operation, which adds handling and increases surface scratch risk.

Gauge profile affects width decisions as much as average thickness. A master roll of PET film with nominal 23 µm thickness may exhibit cross-web variation of ±1.5 µm. When slitting into 100 mm rolls, a narrow high-thickness band can produce a local diameter increase of 1–2%, generating hard bands under rewind pressure. The converter observes this as a washboard surface and intermittent register drift. Slitting patterns are therefore adjusted so that edge trim removes the outermost gauge bands, and the remaining roll widths are aligned with flat zones from the online thickness scanner. Average thickness is verified according to DIN EN ISO 4593, but the converter’s width decision also uses transverse thickness data from the film manufacturer. Published data for this specific configuration is limited; the converter relies on the supplier’s thickness profile chart and slit-position log.

Why Does Treatment-Side Marking Fail at Unwind?

Treatment-side marking often fails because the corona discharge side is identified at the primary film line but the mark is not designed to survive conversion. A corona-treated PET surface typically measures 50–56 dyn/cm under ASTM D2578, while the reverse side may read below 42 dyn/cm. Converters use a dyne pen or spread of mixed wettability solutions from 30 dyn/cm to 56 dyn/cm to verify the side. The information is only useful if the roll-build orientation is stable. When a roll is wound with the treated side inward, the functional surface is protected from dust but can become blocked or contaminated by backside slip additive. When the treated side is outward, the corona state may decay through contact with ambient moisture and temperature. The converter must not infer treatment side from visual haze alone; PET film has low haze and both sides can appear identical under plant lighting. Side-marking is often performed with a water-soluble ink dot on the core or the roll edge, but core labels detach during handling. The most robust approach is to specify the treatment side in the purchase order, record the unwind orientation on the batch traveller, and verify the dyne level at the first idle station after unwind. If the treated side is misidentified, the defect appears as reticulation, pinholing, or adhesive transfer failure after lamination; peel strength drops below specification and the failure surface shows no PET fibre tear.

Corona discharge in air oxidizes the PET surface, creating carboxyl, hydroxyl, and carbonyl groups; this raises wetting tension but leaves low-molecular-weight oxidized material that can bloom or migrate. Converters observe the result as dyne decay from 54 dyn/cm to 46 dyn/cm over weeks in warm storage. Chemical priming is specified when the converter needs a more stable surface for solvent-based inks, but the primer layer alters heat-seal behaviour and may require retesting under EU Regulation (EU) No 10/2011 if the final structure is food-contact. Wetting tension is a necessary but not sufficient measurement; it does not identify which side was treated if both sides were exposed to corona at different levels. The treatment side must be confirmed by a differential wetting check on both surfaces.

Evidence from converter slitting logs indicates that core-related defects are disproportionately associated with unstable rewind tension, not merely core diameter. A 76.2 mm fibreboard core with a wall thickness of 12.7 mm supports slit rolls up to about 40 kg; above that mass, the converter often specifies a 152.4 mm core or a thicker wall. The core must fit the unwind shaft without radial play. Core ID tolerance for fibreboard cores is commonly held to -0/+0.3 mm; a larger clearance causes chuck slip at speeds above 150 m/min. Crushed or out-of-round cores generate cyclic tension variation that appears as intermittent registration drift on a flexographic press. Telescoping—the axial displacement of edge layers—occurs when the rewind taper tension is too low at the final wrap or when the film has low internal friction. For PET film, coefficient of friction is often modified with slip additives to a target of 0.25–0.45 under ISO 8295:1995; extremely low slip can cause the wound roll to shift during transport. Converters check the roll by pressing the end face and observing layer movement. A stable roll has no edge displacement above 1 mm. Core material is also important: paper cores are acceptable for typical packaging film, but wide master rolls above 2,000 mm with high unwind tension may require steel or aluminium cores to prevent barrel collapse.

When Roll Width Tolerances Collide with Rewind Tension

Roll width tolerance interacts with rewind tension in a way that is frequently underestimated. A slit roll with a width deviation of -0.5 mm will have less contact area against the rewind pressure roller, changing the effective nip pressure and producing a harder roll at the edges. If the slitter is running at 300 m/min and edge weave exceeds 0.5 mm, the film will enter the rewind with a slight lateral oscillation. The result is not a formal width defect but a track density difference: edge hardness rises, core compression occurs, and the treated side may be scratched by the pressure roller. On a two-drum winder, the roll weight is distributed over the two support drums, and narrow rolls below 50 mm can develop star defects because the effective contact area is too small. The converter compensates by reducing rewind tension to 0.5–1.0 N/mm of web width and by using lay-on rollers with closed-loop pressure control. Width measurements should be taken after the rewind reaches 60% of final diameter, because the roll edges relax and may show 0.2–0.4 mm width gain in thick film above 100 µm. These variations are within typical specification but may be sufficient to shift side lay on downstream printing presses. The functional control is therefore not a single width number but a width-versus-rewind-hardness profile recorded for each slit position.

At the laminating nip, a converter’s first confirmation of treatment side is often the wet-out behaviour of the adhesive. A solventless polyurethane adhesive applied to untreated PET displays poor wetting, visible beading, and low coating weight uniformity. On the treated side, the adhesive should wet the web within a contact angle range of 20–40° when measured by ASTM D5946-17. The corresponding wetting tension by ASTM D2578 is usually 50 dyn/cm or higher. The incoming specification may allow 48–56 dyn/cm, but the converter must account for treatment decay during storage. Polyethylene terephthalate film stored at 30 °C and 60% RH can lose 2–6 dyn/cm in the first six months depending on the corona density and additive bloom. The converter therefore sets a minimum 42–44 dyn/cm at the point of coating; if the reading is below that, the roll is re-treated on a corona treater or downgraded. Treatment side must be verified at both ends of each slit roll because additive migration and contact with packaging materials can create nonuniform surface energy. The methods in Table 2 provide independent checks; dyne solutions are fast but sensitive to contamination and operator technique, while contact angle measurement is slower but more reproducible for PET film because it is not influenced by surface roughness in the same way.

Core Plug, Chuck, and Brake Interface Limits

The core-to-chuck interface defines the maximum allowable unwind speed and roll weight. A core with internal diameter of 152.4 mm can accept a larger expanding chuck, but the chuck must not expand more than 0.25 mm beyond the core ID or the core end can split. In high-speed laminating at 300 m/min, the unwind brake must maintain web tension without generating core-end heat; paper cores degrade above 100 °C under prolonged friction. Converters specify steel-core adapters or reducer rings when the roll mass exceeds 500 kg. Brake torque is set from the unwinding roll diameter, not the core diameter; a 1,000 mm diameter roll on a 152.4 mm core requires different torque than a 600 mm diameter roll. If the core ID is too large, the chuck slips and the web tension drops, producing slack and misregistration; if the ID is too small, inserting the shaft damages the core and creates particulate contamination. Core ends should be free of crush damage above 1 mm depth. The converter inspects core roundness using a two-point internal gauge before loading; a core that is more than 0.5 mm out of round will produce a once-per-revolution tension cycle visible as repeating print length variation. These interface limits are not commonly tested in a laboratory; they are observed as time-series tension plots on the unwind drive.

In metallization, the treatment side specification carries an additional constraint. The corona-treated side must be the outer surface if the converter is vacuum metallizing in a roll-to-roll chamber at pressures below 10⁻³ Pa. If the wrong side is placed toward the evaporation source, aluminium adhesion is poor and the film shows streak defects. For printing, the treatment side must face the first printing station; reverse printing through the film requires treatment on the opposite surface. A two-side treated PET film is sometimes specified for multi-layer structures, but the second-side treatment can alter coefficient of friction and blocking behaviour. Converters control this by specifying one-side treatment for most lamination and by requiring a slip-modified backside with 0.3–0.4 coefficient of friction under ISO 8295:1995. The roll width and core diameter must be compatible with this surface orientation because reversing a roll at a secondary unwind can transfer slip additive to the treated surface. When purchasing PET film under FDA 21 CFR 177.1630 or EU Regulation (EU) No 10/2011 for food contact, the converter includes side and surface specifications in the compliance certificate. Published data for specific combinations of roll width, core ID, and treatment orientation is limited; converter specifications therefore rely on incoming inspection records rather than a single global standard.