Injection Moulding Polycarbonate: Drying, Melt Viscosity and the Cycle That Holds

Polycarbonate (PC) is an amorphous condensation polymer with a glass transition temperature near 147°C and a practical melt-processing band of 280°C to 320°C. On injection moulding lines the interaction between residual pellet moisture, melt viscosity and the pack/hold segment dictates whether a part retains its specification. The material absorbs moisture from ambient air and hydrolyses through carbonate-link cleavage at melt temperatures; it also exhibits a temperature-sensitive melt viscosity whose control requires a narrow process window.

Production dryers for PC are sized for the worst expected relative humidity in the moulding hall, not for the average day. A closed-loop desiccant dryer with a return-air dew point between -40°C and -30°C is standard when the target pellet moisture is 0.02% or lower. Pellets removed from a sealed container and held in an open hopper at 23°C and 50% relative humidity can approach equilibrium moisture contents of 0.15% to 0.20%; only a fraction of that value is permissible before melt-phase hydrolysis becomes measurable. Drying temperature is kept at 120°C for 4 h for virgin PC in a properly functioning desiccant dryer, while regrind and material from open containers may require 6 h to 8 h. The drying hopper air temperature is not a substitute for dew-point control: if the desiccant bed is saturated and return-air dew point rises above -20°C, the drying rate falls sharply even at 120°C. Overdrying has a boundary as well: drying PC above 130°C can cause pellet surface oxidation and yellowing, and residence beyond 8 h at 120°C is not recommended unless the hopper is padded with dry air or nitrogen.

Residual moisture in PC is not uniformly distributed through the pellet; surface moisture desorbs quickly, while centre moisture migrates by diffusion. Drying therefore requires residence time as well as dew point. A high air temperature cannot reduce the required time below the diffusion limit without risking pellet fusion or oxidative yellowing. Dryer audits that measure only inlet temperature miss partial desiccant-bed failure because the outlet humidity may rise while the temperature stays constant.

When residual moisture enters the melt stream: hydrolysis kinetics and dryer limits

Hydrolysis of the carbonate linkage proceeds by random chain scission. The melt flow rate measured to ISO 1133-1:2022 at 300°C and 1.2 kg therefore becomes both a lot-release and a process-degradation check. A virgin PC lot with an MFR of 10 g/10 min can show an upward shift of 2–3 g/10 min after repeated exposure to humid air; a shift of 5 g/10 min or more is frequently associated with a measurable loss in notched impact strength when tested to ISO 179-1/1eA. The issue is not limited to surface moisture: PC pellets can retain internal moisture after poor drying, and the pressure drop across the screw compression zone does not volatilise water at 300°C; it disperses superheated vapour that hydrolyses the melt and produces splay, silver streaks and bubbles in thick sections. Vent zones on a two-stage screw remove trapped air and volatile surface moisture but do not reduce water content below the hydrolysis threshold.

Hot-runner systems intensify the requirement: melt residence time in a heated manifold can exceed 10 min during interruptions, and with residual moisture at 0.03% the cumulative chain-scission effect may appear as viscosity reduction and brown streaking in subsequent shots. Karl Fischer titration to ISO 15512 is the reference method for testing PC pellet moisture. A sample mass of 1 g to 2 g is typical, and the method reports water in weight percent. A moisture specification of 0.02% is equivalent to 200 ppm; however, Karl Fischer titration detects total water, including tightly bound water that is not removed by desiccant drying. Processors therefore combine the titration value with an MFR check after a dry-heat exposure to confirm that the water is not hydrolytically active.

What are the melt viscosity boundaries for repeatable filling?

Polycarbonate melts are pseudoplastic but less shear-thinning than typical semi-crystalline olefins. Capillary rheometry to ISO 11443 at 300°C shows apparent viscosities commonly between 200 Pa·s and 600 Pa·s at 1000 s^-1 for medium-viscosity injection and extrusion grades, while high-flow grades used for thin-wall medical and optical parts may fall below 150 Pa·s at the same shear rate. At lower shear rates in hot-runner manifolds, viscosity can exceed 800 Pa·s to 1200 Pa·s, so manifold temperature uniformity becomes critical. A 10°C increase in melt temperature can reduce viscosity by about 10% to 20% within the 280°C to 320°C processing band, but the coefficient varies with molecular weight and shear rate. Above 320°C thermal degradation and yellowing dominate; below 280°C viscosity produces unfilled thin sections and elevated injection-pressure demand. For optical grades the melt-temperature window is frequently held to ±5°C around the set point because larger deviations alter shear heating, flow-front temperature and final part birefringence.

Injection speed is set to keep the flow front advancing because PC develops a solidified skin quickly against mould walls. Thin-wall parts with nominal wall thickness below 1.5 mm often require fill times under 0.5 s, while thick-wall light guides may use slowed filling to preserve optical transmission. Screw recovery back pressure between 5 MPa and 15 MPa is typical on reciprocating-screw machines to compact the melt and assist pigment dispersion; sustained back pressure above 20 MPa raises melt temperature and hydrolysis risk without improving homogenisation. Solid density is approximately 1.20 g/cm³ to ISO 1183-1, and melt density at 300°C is roughly 1.08 g/cm³ to 1.10 g/cm³; this difference creates the volumetric shrinkage that must be managed in the hold phase.

Processors using hot runner systems often compensate for viscosity by raising manifold temperatures, but local overheating can generate black specks and brown striations. A manifold temperature difference of 10°C across the drops can be sufficient to cause noticeable colour variation in transparent parts. Thermocouple placement in the manifold should be verified with a calibrated contact probe because the displayed value may lag the actual melt temperature by several seconds.

Hold pressure, gate freeze-off and the cycle that holds

The injection moulding cycle for PC departs from semi-crystalline practice in the hold phase. Amorphous PC does not generate a crystallisation plateau, so the pack/hold stage is governed by gate freeze-off, pressure-volume-temperature response and local shrinkage compensation. Transfer from velocity control to pressure control is set at 95% to 99% of volumetric fill; transfer too late can spike cavity pressure and damage vents. Pack pressure is usually maintained at 50% to 70% of the peak injection pressure, but the required value is best determined by cavity-pressure sensors because PC's high melt viscosity creates a large pressure gradient between the injection unit and the last-filled region. Hold time is continued until gate seal; for PC with a gate diameter below 1.0 mm, gate freeze time may be 1 s to 3 s, while a 3 mm gate and 3 mm wall can require 8 s to 15 s of hold. Insufficient hold time produces sink marks, vacuum voids at thick-to-thin transitions, and optical birefringence in transparent parts; excessive hold time transmits pressure to already-filled areas and can increase residual stress and ejection force.

Cavity-pressure sensors are preferred over hydraulic-pressure-based estimates because intensification ratio alone does not capture the pressure drop across the gate and runner. A sensor placed in the last-filled region should record a peak cavity pressure of 30 MPa to 60 MPa for many technical PC components; thick-wall optical parts may require 40 MPa to 70 MPa. Gate seal time is confirmed by a flat pressure trace after the hold phase or by incremental hold-time trials with a 0.5 s step. If cavity-pressure sensors record a post-filling pressure drop of more than 25% before gate freeze, the cycle that holds is too short or the gate is too large to freeze at the intended time.

Mould surface temperature is held between 70°C and 110°C for general-purpose PC. Optical and low-stress applications use the upper end, sometimes up to 120°C, to reduce frozen-in orientation and improve replication of microstructured surfaces. Mould temperatures below 70°C reduce cycle time but increase residual stress near the gate; notched impact testing per ISO 179-1/1eA can reveal this process-induced shift before dimensional checks do. Mould shrinkage, measured to ISO 294-4, typically falls between 0.5% and 0.7% for unfilled PC, but the value depends on wall thickness, melt temperature, mould temperature and hold pressure. Vent depths for PC are typically 0.02 mm to 0.04 mm; inadequate venting at the end of fill can produce gas burns and short shots even when the machine pressure reading remains stable.

Cooling time is usually the longest phase. For a 3 mm wall, cooling time is often 20 s to 60 s depending on mould temperature. Ejection is permissible when the part core reaches below 130°C, but for dimensional stability the surface temperature at ejection should be below 110°C. Ejector speed and pattern are set to avoid deformation because PC retains low modulus near its glass transition. Tie-bar strain and mould deflection can also influence the cycle that holds. On large multi-cavity tools, uneven clamp pressure distribution causes the parting line to breathe, especially at cavity pressures above 60 MPa. This breathing can produce local flash and reduce the pressure available for shrinkage compensation in the centre of the tool. The clamp force should be set on the basis of measured cavity pressure and projected area, not a historical value, and the tool should be checked for parallelism under full clamp force.

Thermal degradation pathways at the 300–320°C processing boundary

Prolonged melt residence time at the high end of the processing band causes random chain scission, discoloration and gas evolution. The Yellowness Index measured to ASTM D1003 is the practical release indicator for transparent grades; yellowing is usually more sensitive than MFR at the early stages of degradation. In hot-runner systems, melt residence time should be kept below 15 min when the melt temperature is 310°C or higher. For critical optical parts, barrel temperature profiling is set with a reverse profile: the rear zones are typically 10°C to 20°C higher than the middle zones, while the nozzle is reduced by 5°C to 10°C to balance shear heating. Published data for the precise degradation onset varies by additive package, and not all grades tolerate the same combination of residence time and moisture.

Gate geometry, wall thickness and pressure drop in practice: PC moulds benefit from full-round or trapezoidal runners because the melt tends to freeze quickly at runner walls; sharp transitions create stagnation zones that require longer hold to pack. Direct edge gates generate lower residual stress compared with pin-point gates in multi-cavity tools, but gate size must account for the pressure required to fill 2.5 mm to 4 mm walls. For thick-section parts above 5 mm, process development uses sequential valve-gate opening to avoid flow hesitation and local overpacking. Clamp force requirements are generally higher than for polypropylene at similar projected area because cavity pressure must be maintained during the hold phase; machines are commonly selected with clamp force of 0.5 t/cm² to 0.8 t/cm² of projected area for moderate flow lengths. Regrind use in PC is limited by molecular weight and contamination: 20 wt% to 30 wt% regrind is typical for non-appearance parts, but each heat history can raise MFR by 1 g/10 min to 3 g/10 min depending on drying, temperature and residence time. For medical applications regulated under ISO 10993 and food-contact applications requiring FDA 21 CFR 177.1580 compliance, regrind is generally restricted to first-generation material from the same production lot.

Post-mould annealing is sometimes used for PC components under high chemical exposure. Annealing at 120°C for 2 h in a circulated-air oven reduces residual stress but can increase dimensional change by releasing frozen-in strain; the cycle that holds must therefore be tuned to minimise stress before post-processing rather than relying on annealing as a corrective step.