Polyethylene terephthalate melt processing is governed by moisture control before plastication and by shear and residence-time restrictions during extrusion. Commercial PET grades for bottle, sheet, film, and strapping are supplied with an intrinsic viscosity commonly between 0.70 dL/g and 0.85 dL/g; the test method is ASTM D4603-18. Residual moisture after desiccant drying must be held below 50 ppm and is normally targeted at 30–50 ppm. At melt temperatures of 270–285 °C, water present in the melt hydrolyses ester linkages and reduces intrinsic viscosity by 0.02–0.05 dL/g per pass, while acetaldehyde concentrations increase in the finished article. Production lines therefore use a crystallizer at 150–170 °C for amorphous PET and a desiccant dryer with a dew point of −40 °C or lower, monitored by a dew-point sensor in the return air.

On a 90 mm single-screw extruder with 30:1 L/D, barrel set points are typically rear 260–270 °C, compression 270–280 °C, metering 275–285 °C, and adaptor/melt pump 270–280 °C. Melt temperature excursions above 300 °C accelerate thermal-oxidative degradation, acetaldehyde formation, and visible gel defects. Screw speed, vacuum level, and melt filtration must therefore be treated as a single system: excessive shear cannot be undone by vacuum venting, and inadequate vacuum cannot be corrected by finer filtration.

A production-scale failure mode observed on sheet and film lines is the slow formation of amber or black specks at the die lip, followed by a pressure drop in the melt pump and a rising vacuum pressure at the vent. This sequence indicates hydrolytic chain scission combined with stagnant melt accumulation. At ambient relative humidity above 60%, pellet handling must be coupled to a closed conveying system with nitrogen purge; open hopper loading can raise residual moisture above 50 ppm during typical multicomponent changeover periods.

What Screw Geometry Restricts Shear Heating and Hydrolytic Chain Scission in Single-Screw Extrusion?

Barrier flight geometry is preferred over a simple three-zone screw because it melts PET without trapping solid bed fragments in the metering channel. Compression ratios are specified from 2.5:1 to 3.0:1; higher compression ratios generate excessive viscous dissipation and melt-temperature rise. The total length is held between 24:1 and 30:1 L/D for general sheet and film, while higher-output lines may extend to 36:1 L/D when the additional length is devoted to decompression and vacuum venting rather than shear. The barrier flight clearance is maintained at 0.25–0.50 mm. A clearance below 0.25 mm raises the shear rate in the melt channel and contributes to acetaldehyde and gel formation; a clearance above 0.50 mm allows unmelted pellets to bypass the barrier and reach the metering section.

For a 90 mm screw, a metering depth of 3.0–4.5 mm is typical for high-IV PET. The feed depth is normally 8.0–9.5 mm to control the compression ratio while maintaining output. Peripheral screw speed is limited to 0.30 m/s; on a 90 mm extruder this corresponds to approximately 64 rpm, although higher speeds are used only with shallow metering depths and short residence-time barrels. Melt temperature rise from shear is measured at the adaptor with an immersion thermocouple, and the difference between the barrel set point and melt temperature should not exceed 5–8 °C for unfilled bottle-grade resin.

The solids-conveying zone is single-flighted with a pitch equal to the screw diameter and extends for 8–10 L/D. The barrier section begins at approximately 50–60% of the total screw length and runs for 6–10 L/D. In a two-stage vented configuration, the decompression zone behind the vent has a depth of 1.5–2.0 times the first-stage metering depth to keep the melt partly filled; the second-stage metering depth is then selected to provide a pumping capacity 10–20% above the first stage. Screw and barrel wear is critical because the relatively low melt viscosity at 280 °C reduces the hydrodynamic film thickness; screws are chrome-plated or coated with chromium carbide to prevent iron-catalysed degradation at the barrel surface.

Low-shear mixing sections are used sparingly. Maddock-style mixers can raise melt temperature by 5–10 °C and are justified only when colour or additive distribution is otherwise unacceptable. For PET, a slotted or pineapple mixer with wide clearances is preferred to a high-shear dispersive mixer. The feed throat is water-cooled at 40–60 °C to prevent pellet bridging; the first barrel zone is set at 260–270 °C to avoid premature melting and compression-driven shear. On a vented two-stage screw, the first-stage metering pressure is typically 80–150 bar, and the melt pump inlet pressure is held at 30–50 bar to avoid over-shearing in the pump gears.

Published data for specific high-output PET extrusion configurations is limited. Table 1 lists representative industrial starting values that must be confirmed by rheological and optical defect measurement on the target production line.

A two-stage vented screw creates a decompression zone before the vent port at an axial position of 20–24 L/D. The first-stage metering section pumps fully molten polymer into a deeper, partially filled channel where pressure falls and volatile substances desorb. The second stage must pump 10–20% more volume than the first stage to maintain a melt seal and to prevent the melt from climbing into the vent port. Vacuum level is maintained at 10–50 mbar absolute using a liquid-ring or rotary-vane pump equipped with a condensate trap. Lower vacuum fails to strip acetaldehyde and water; higher vacuum can carry oligomer fog and low-viscosity degraded polymer into the vent line, where it carbonizes and eventually re-enters the melt stream as black specks.

Vacuum pump capacity is selected for the volatiles load from the resin and regrind; a starting point for sheet lines is 30–50 m³/h per 100 kg/h of PET throughput, but this varies with the target acetaldehyde specification. Vacuum lines are sloped to a chilled trap maintained at 5–10 °C to condense glycol and oligomer sublimates. At start-up, the vent is kept at atmospheric pressure until a stable melt seal is observed; vacuum is then applied gradually. Applying vacuum too early pulls unmelted particles and low-viscosity melt into the vent, leading to immediate fouling.

Vent flooding occurs when the fill fraction at the vent port exceeds 0.60–0.70, when the second-stage pumping capacity is too low, or when the vacuum line is fouled with condensed glycol and oligomers. On a sheet line, vent flooding appears as cyclic pressure at the melt pump, material pumping from the vent, and a rapid increase in gel count at the die. The immediate response is to reduce screw speed by 10–20% and increase the downstream barrel temperature by 5 °C; however, repeated flooding indicates that the screw geometry is mismatched to the melt viscosity of the PET grade being processed.

Vacuum venting removes water, acetaldehyde, ethylene glycol, and short-chain oligomers, but it does not remove crosslinked gel particles. Gel control therefore requires oxygen exclusion, low residence-time flow paths, and mechanical filtration. A hopper nitrogen purge of 2–5 L/min and an atmospheric vent upstream of the vacuum port reduce oxidative degradation in the first-stage metering zone. Continuous screen changers with 40/60/100 mesh packs remove particles above 40–80 µm. A gear pump stabilizes melt pressure at 180–250 bar and prevents pressure surges that can break filter cakes and release agglomerated gels downstream.

When Hydrolytic Degradation Outpaces Gel Formation from Crosslinking

Gel defects in PET extrudate are broadly divided into hydrolytic specks and thermal-oxidative crosslinked particles. Hydrolytic specks form when residual moisture above 50 ppm attacks ester linkages, producing low molecular weight fractions that crystallize and appear as haze or white specks in sheet. Thermal-oxidative crosslinked particles form in stagnant zones, hot spots, or high-shear regions where the melt temperature exceeds 290 °C for even short periods. Both mechanisms can coexist because water and oxygen ingress at the feed system, vent port, or regrind stream accelerate chain scission and crosslinking simultaneously.

Control begins with melt temperature and residence-time distribution. Barrel zones above 285 °C should be confined to the final 1–2 L/D before the melt pump. Average melt residence time in a 90 mm extruder is held below 3–5 min by matching screw speed to output; stagnant zones in adaptors, screen changers, and melt pump housings must be streamlined and dismantled during scheduled maintenance. Contamination from polyvinyl chloride or acetal-based purge compounds is excluded because acid degradation products accelerate hydrolysis and gel formation. For high-IV PET, the melt pump inlet pressure is maintained below 50 bar to limit shear in the pump gears, and the differential pressure across the screen changer is kept below 50–80 bar to prevent gel cake re-release.

Mechanical filtration at the breaker plate removes gel particles and dispersed foreign matter. A screen pack of 40/60/100 mesh provides a nominal filtration rating of approximately 40–80 µm, but the actual cut point depends on the depth of the filter cake and the melt viscosity. Continuous screen changers are preferred because pressure spikes during manual screen changes can exceed 20 bar and force agglomerated gels through the filter. Reclaim streams containing PET regrind are limited to 10–30 wt% for film and sheet because each heat history increases gel precursors and carbonyl species; higher regrind fractions require finer filtration and more aggressive vacuum venting to maintain equivalent optical defect limits.

The interaction between screw geometry, vacuum venting, and gel control is direct. A barrier screw that leaves unmelted pellets creates hydrolytic gel precursors in the final product; a vent that floods introduces carbonized deposits into the melt; a screen changer that generates pressure surges shears the polymer and creates fresh gel particles. Process limits are therefore monitored as a combined set of melt temperature, vent vacuum, differential pressure, and intrinsic viscosity loss. A stable process for bottle-grade PET typically maintains a per-pass intrinsic viscosity loss below 0.03 dL/g per ASTM D4603-18 and acetaldehyde below the target bottle or sheet specification measured per ASTM F2013-10.