Incoming bottle flake sourced from curbside bales is not a homogeneous rheological feedstock. A 30 mm corotating intermeshing twin-screw extruder processing recycled polyethylene terephthalate must contend with intrinsic viscosity spread, bulk density variation, residual label adhesive, polyvinyl chloride contamination, and fluctuating carboxyl end-group concentrations before the first melting zone is reached. The extruder is normally starve-fed from a gravimetric feeder, and the feed stream can exhibit bulk density between 0.28 kg/L and 0.42 kg/L after wash and hot-drying. Intrinsic viscosity for bottle-grade rPET flake typically falls between 0.70 dL/g and 0.80 dL/g when measured per ASTM D4603-18, but batch-to-batch shifts of 0.05 dL/g or more are common when post-consumer bales are blended. The same flake may contain 2000 ppm to 5000 ppm of surface moisture after ambient storage, and as ambient relative humidity exceeds 60%, pre-drying at 160–170 °C for 4–6 h with a desiccant-bed dew point of -40 °C or lower is required to drive moisture below 50 ppm before the main feed throat. Karl Fischer moisture testing per ASTM D6869-17 is used to confirm the dried flake moisture content, because residual water above 80 ppm begins to hydrolyze the ester linkages in the early kneading blocks and can reduce pellet intrinsic viscosity by 0.03–0.06 dL/g before the melt reaches the vacuum vent. The blending window is therefore set by the interaction of flake drying history, feeder accuracy, and the residence time required to disperse residual polyolefin cap material that remains from insufficient sink-float separation.

What Limits Melt Temperature Across Barrel Zones 2 to 5 in a 30 mm rPET Line?

Melt temperature in recycled bottle flake is bounded on the low side by incomplete melting and on the high side by thermal degradation. For a 30 mm corotating twin-screw extruder with a 40:1 length-to-diameter ratio, the first barrel zone is commonly set at 240–260 °C to initiate fusion without prematurely compacting flake into an unmelted plug. Barrel zones 2 through 5 are maintained at 260–280 °C, while the die adapter and melt pump zones are limited to 275–285 °C. The measured melt temperature at the die should remain between 270 °C and 285 °C; excursions above 295 °C produce measurable acetaldehyde formation and yellow shift, while sustained operation above 300 °C causes rapid random chain scission and a loss of intrinsic viscosity exceeding 0.05 dL/g in a single pass. On the low-temperature side, a melt temperature below 255 °C is often associated with high motor torque, poor melt sealing at the vacuum vent, and insufficient dispersion of solid chain extender particles. The narrow processing window is driven by the combination of high screw speed, viscous dissipation, and the low thermal conductivity of the partially crystalline flake. Melt temperature is not equal to barrel setpoint; localized melt temperature in a 30 mm machine can exceed the barrel wall temperature by 10–20 °C under high specific torque conditions. The acceptable operating corridor is therefore narrower than the barrel setpoint range suggests, and any increase in throughput that raises specific mechanical energy beyond 0.32 kWh/kg must be compensated by reducing screw speed or increasing barrel temperature only within the degradation boundary.

The screw configuration that provides a stable melt seal at L/D 28 while still supporting additive dispersion is not a single all-purpose design. A typical 30 mm corotating TSE used for rPET compounding is assembled from right-handed conveying elements with pitch lengths from 30 mm to 20 mm, followed by neutral and left-handed kneading blocks located after the side-stuffer port. The outer diameter to inner diameter ratio of the screw profile is usually 1.55, giving a deep flight channel that limits shear heating but reduces dispersive mixing relative to a 1.66 ratio. A machine with a specific torque rating of 8.5 Nm/cm³ operates conservatively for bottle flake, while a 11.3 Nm/cm³ gearbox permits higher throughput at the same screw speed before the torque limit is reached. At a screw speed of 400 rpm, a 30 mm high-torque machine typically delivers a mean residence time between 35 s and 55 s, but the tail of the residence time distribution can extend to 90–120 s when low shear zones exist in the melt seal area. This tail contributes to stagnation and gel formation if reactive additives accumulate in the apex region of the kneading blocks. Screw configurations with three to four left-handed kneading blocks distribute mixing energy more uniformly, but they also increase melt seal pressure and may force material degradation at the vent if the seal becomes too restrictive. The selection of screw elements is therefore a compromise between reaching a vent-stable melt seal pressure of 10–25 bar and limiting the maximum local melt temperature to 290 °C during side-stream addition.

If a Side-Stuffer Is Placed at L/D 24 and the First Vacuum Vent at L/D 32, Additive Distribution Is Set by the Melt Seal at L/D 28

Additive addition through a side-stuffer is governed by the pressure drop across the melt seal and the particle size of the solid component. In a 30 mm machine, side-stuffing is commonly performed with a co-rotating auxiliary screw of 16–20 mm diameter operating at 50–150 rpm. The side-stuffer must overcome the local melt pressure in the main barrel, which should not exceed 8–15 bar at the addition port if the side-stuffer is to feed consistently. If the main screw speed exceeds 600 rpm and the side-stuffer feed rate exceeds 4 kg/h, low-bulk-density additives can short-circuit along the screw channel toward the vacuum vent unless a left-handed element at L/D 28 generates an upstream melt seal. The same melt seal increases residence time for dispersive mixing but also creates a zone of elevated pressure that compresses solid additives into the melt surface. Published data for this specific 30 mm configuration is limited, and machine supplier start-up guidance typically relies on pressure transducers placed at L/D 25 and L/D 30 to verify that the melt seal pressure remains between 10 bar and 25 bar. When the seal pressure drops below 8 bar, the vacuum port may flood; when it exceeds 30 bar, the side-stuffer motor stalls or the melt temperature rises sharply. The blending window is therefore defined by the screw geometry between L/D 24 and L/D 32, not by the feeder calibration alone.

Reactive chain extension on a 30 mm TSE is typically performed with solid branched chain extenders such as pyromellitic dianhydride or tetraglycidyl diamino diphenylmethane. Addition levels are kept between 0.15 wt% and 0.40 wt% relative to rPET, with the lower end used when the incoming flake has an intrinsic viscosity above 0.75 dL/g and the upper end reserved for heavily degraded post-consumer material with an intrinsic viscosity below 0.65 dL/g. The chain extender must be fed as a free-flowing powder under a nitrogen blanket to prevent moisture pickup, because anhydride and epoxy functionalities hydrolyze before reaching the melt if exposed to humid air above 60% relative humidity. The reaction itself competes with thermal degradation, and a residence time of 30–60 s at 270–280 °C is usually sufficient for chain branching without forming gelled regions. Excess chain extender above 0.5 wt% creates localized network formation, increasing head pressure and reducing melt pump suction stability. In addition, amine-based heat stabilizers must be avoided because they react with epoxide or anhydride groups prematurely and shift the stoichiometric balance away from chain extension. The operational boundary for reactive rPET blending is therefore set by additive moisture, main-feed moisture, and the ratio of anhydride or epoxide functionality to available carboxyl end groups.

Vacuum Venting, Acetaldehyde Stripping and Polymer Melt Seal Stability

At the vacuum vent, the melt must be sufficiently fused to present a continuous surface for stripping but not so fluid that it migrates into the vent port. The absolute pressure in the vent dome is held between 2 kPa and 10 kPa, which corresponds to 15–75 mmHg, to remove acetaldehyde, water, glycols, and low-molecular-weight volatiles. A stable melt seal upstream of the vent is required because any break in the seal allows unfilled screw channels to carry molten polymer directly into the vacuum line, where it can solidify and block the condenser. Flooded vents are a common failure mode when the screw speed is raised too quickly or when the throughput is increased without a corresponding increase in the seal pressure. Acetaldehyde stripping efficiency is strongly temperature dependent, and the vent temperature should remain above 260 °C to keep the melt surface at low enough viscosity for diffusion of the aldehyde from the melt. Vacuum pump capacity on a 30 mm line is usually between 20 m³/h and 60 m³/h, with a condensate trap to prevent low-volatile waxes from reaching the oil-lubricated pump. For bottle-grade rPET, the target residual acetaldehyde is below 1.0 ppm when measured by headspace gas chromatography per ASTM F2013-10(2021), while sheet and thermoforming grades may accept up to 2.5 ppm. The venting window cannot be established by vacuum pressure alone because the melt seal stability and the concentration of dissolved volatiles in the incoming flake define the practical lower limit.

Filtration of a rPET melt on a 30 mm line is performed through a duplex slide-plate screen changer equipped with a breaker plate and a screen pack typically assembled from 50/100/50 mesh layers. The filtration area on a 30 mm machine is usually between 10 cm² and 20 cm², and the head pressure upstream of the screen changer is maintained between 20 bar and 70 bar. A pressure drop exceeding 40 bar across the screen pack indicates the presence of aluminum flake, paper fiber, or unmelted PVC contamination in the melt stream; automatic screen change is triggered when the upstream pressure reaches the 70–80 bar range to prevent excessive shear heating. The melt is then metered through a gear pump to damp pressure fluctuations before entering an underwater pelletizer. Die plate temperature is held at 280–295 °C, and the water temperature is set between 35 °C and 50 °C to control pellet crystallinity and prevent sticking in the spin dryer. If the pelletizer water temperature drops below 30 °C, the quenched pellets may have an amorphous skin that fuses in storage; if it exceeds 60 °C, pellet agglomeration increases. The throughput limit for pelletizing on a 30 mm line is usually governed by the die hole count and not by the screw capacity, because high-shear die flow can raise the melt temperature by 5–10 °C just before quenching.

Intrinsic Viscosity Retention From Flake to Pellet Is a Linear Function of Moisture, Temperature and Residence Time

Retention of molecular weight through the 30 mm TSE is measured by comparing flake intrinsic viscosity to pellet intrinsic viscosity per ASTM D4603-18. A well-operated line processing 100% bottle flake at 285 °C melt temperature and 0.25 kWh/kg specific mechanical energy typically retains 0.03–0.07 dL/g of the original intrinsic viscosity, while higher thermal histories can increase the loss to 0.10 dL/g or more. The MVR measured per ISO 1133-1:2022 at 280 °C and 2.16 kg rises as the intrinsic viscosity falls, and a shift from 0.75 dL/g to 0.70 dL/g can increase MVR by 5–10 cm³/10 min. Moisture is the strongest independent accelerator of molecular weight loss; at 100 ppm water in the melt, hydrolysis occurs alongside thermal chain scission, and the combined effect reduces intrinsic viscosity by 0.08–0.12 dL/g even at the same barrel temperatures. The linear relationship between moisture, residence time, and intrinsic viscosity loss is the reason that drying and venting are treated as inseparable, not as separate unit operations. Chain extension can restore 0.04–0.08 dL/g when reactive additives are properly dispersed, but it cannot fully compensate for severe hydrolysis because the carboxyl end groups generated by water attack compete with the chain extender and produce branched low-molecular-weight fractions rather than linear chain extension.

Pellet color and haze cannot be interpreted independently of crystallinity and additive scattering. Bottle flake can enter the 30 mm extruder with CIE L* values from 65 to 75 and b* values from -1 to 3 as determined per ASTM D6290-19. The compounding operation itself can shift b* by 1.5–2.5 units if the melt temperature is allowed to exceed 290 °C, and the same thermal history accelerates the production of oxidized species that increase haze in downstream injection-molded preforms. Haze is measured on molded plaques per ASTM D1003-21, and high values are frequently traced not to the base rPET but to incompatible polyolefin contamination from cap material or to undispersed impact modifier domains smaller than the wavelength of visible light. The processing window for color-critical rPET therefore requires strict control of melt temperature, vent vacuum, and screen changer pressure, because yellowing and haze cannot be corrected by pellet blending alone once the molecular weight has started to drop. Food-contact suitability for the compounded pellets is not established by processing conditions alone; compliance is assessed under EU Regulation 2022/1616 or applicable national food-contact rules, and the extrusion line must be validated for decontamination efficiency rather than relying solely on the intrinsic viscosity and acetaldehyde numbers.