The First Shot Came Out Brittle: A Polycarbonate Drying Autopsy

The first shot came out brittle from a 1300 kN servo-electric injection molding machine processing a clear polycarbonate grade with a melt volume-flow rate of 12 cm³/10 min at 300 °C and 1.2 kg per ISO 1133-1:2022. The component fractured at the tunnel gate before full ejection while the runner remained intact. The same material lot produced dimensionally stable parts after 15 cycles, indicating that the defect was not a bulk resin-lot variation but a start-up transient. Polycarbonate equilibrates at approximately 0.15 wt% moisture at 23 °C and 50 % RH and reaches approximately 0.35 wt% at saturation per ISO 62:2008. Dried pellets can regain enough surface moisture to exceed the molding limit within 2 h of ambient exposure. The first shot is therefore an early detector of dryer start-up error, feed-throat moisture uptake, and hopper dead-volume.

What does the first-shot failure signature reveal about residual moisture distribution?

Brittle first-shot polycarbonate parts frequently show transverse crack initiation at the gate, silver streaks through the sprue, and a finely blistered surface in the last-filled region. These features overlap with cold-mold and cold-runner artifacts. Hydrolysis of the carbonate linkage consumes water and releases carbon dioxide; at melt temperatures from 280 °C to 320 °C, that reaction converts local water pockets into gas-rich defects. The gate fracture in a moisture-degraded first shot typically corresponds to localized molecular weight loss at the surface of the shot, where the hot screw-root film had the longest contact with moisture-laden polymer. Size-exclusion chromatography of a failed first shot can resolve a 10 % to 20 % reduction in weight-average molecular weight relative to a dried-molded control, while differential scanning calorimetry shows no change in the glass transition region near 149 °C unless severe degradation has occurred.

In a humid plant at 28 °C and 65 % RH, the ambient dew point is approximately 21 °C. If feed-throat cooling water is set to 15 °C, the screw shank and feed-throat walls can fall below the wet-bulb temperature and condense water directly onto dried pellets and the screw root. The first shot then contains a water-rich surface layer that is not detected by a hopper-discharge moisture reading. Vacuum receivers should be purged with dry air at -40 °C dew point, and feed-throat jackets should be supplied with water above the ambient dew point, commonly 25 °C to 30 °C for tropical plant conditions. This is an equipment-level correction that precedes any barrel-temperature or back-pressure adjustment.

Production dryers are frequently started simultaneously with hopper loading, so the controller displays the correct setpoint of 120 °C while the material at the discharge has seen only 30 min of transient heating. In a 200 kg/h desiccant-wheel dryer with an inlet-air dew point of -40 °C, the first pellets delivered to the feed throat can still contain 0.03 wt% to 0.05 wt% moisture because the required 4 h residence time has not elapsed. The failure then disappears after a dozen shots, leading to an incorrect conclusion that the dryer was adequate from the start. The first-shot autopsy needs a time-series record: hopper discharge temperature, return-air temperature, and dew point from the first minute of operation. Without that record, the brittle first shot may be misclassified as a cold-mold problem and the drying transient remains uncorrected.

Desiccant dryer dew point, airflow rate, and return-air temperature boundaries

For polycarbonate, process air should be delivered at 120 °C with a pressure dew point of -40 °C or lower and an airflow rate of approximately 3.7 m³/h per kg/h of polymer throughput. A 50 kg/h line therefore requires near 185 m³/h of dry-air flow. The desiccant-wheel regeneration loop should operate at 180 °C to 220 °C and the cooling leg should return the regenerated bed to below 40 °C before changeover. Return-air temperature is a practical field indicator; for polycarbonate, a value below 80 °C during steady-state drying usually indicates an under-loaded hopper, excessive airflow, or a wet pellet burden. Dew-point sensors should be located after the desiccant bed and after the return-air filter, with a machine start interlock at -30 °C.

Moisture verification should use a heated Karl Fischer vaporizer at 160 °C to 180 °C, because halogen moisture balances can under-report hydrogen-bonded water in the carbonate matrix. Sampling at the hopper discharge alone is insufficient; the sample should be extracted at the machine feed throat after vacuum transfer through plant air. At 70 % RH, a throat sample can increase from 0.018 wt% to 0.045 wt% in 20 min if the feed throat is purged with compressed air rather than dry process air. The screw-root polymer film retains that moisture until the barrel heat drives the hydrolysis reaction, which is why the first shot rather than the hopper sample reveals the true start-up condition.

The barrel temperature profile for a medium-viscosity polycarbonate is usually set from 280 °C near the feed zone to 300 °C at the metering zone and nozzle, with a flat or slightly decreasing nozzle temperature to minimize drool. The hydrolysis reaction is strongly temperature-dependent; a 10 °C increase in melt temperature above 300 °C can double the rate of molecular weight loss in wet resin. Shear heating from high screw rotation and back pressure adds another 5 °C to 15 °C to the melt. The first-shot autopsy should therefore compare the shot-to-shot melt temperature log with the hopper moisture data. If the first shot was processed at 305 °C because of a high back-pressure setting, while steady-state shots ran at 293 °C, the temperature transient amplifies the hydrolysis penalty from residual moisture. This interaction explains why a wet first shot can fail even when the same pellet moisture would not fail a warm, steady-state shot.

If regrind content rises above 30 % by weight without a dryer residence-time correction

Regrind from rejected first-shot parts has higher surface area, lower bulk density, and a greater moisture adsorption rate than virgin pellet. When the regrind fraction exceeds 30 wt%, the hopper residence time at 120 °C should be extended to 5 h to 6 h. Fines below 0.5 mm create hopper bridging and flow channels, producing a bimodal moisture distribution where virgin-rich zones near the hopper wall dry to 0.015 wt% while regrind-rich core zones remain near 0.05 wt%. Injection pressure and screw recovery time shift by 10 % to 15 % because bulk density changes, not because the polymer melt viscosity has degraded. A first-shot failure in a regrind-containing lot is therefore as likely to be a hopper mass-flow problem as a dryer problem, and the corrective action should begin with granulate screening and a residence-time audit.

The cold mold and the cold slug in the nozzle can also produce a brittle first shot from adequately dried polycarbonate. Polycarbonate requires mold surface temperatures between 80 °C and 120 °C to limit frozen-in stress and permit molecular relaxation. A first shot entering a mold at 35 °C can show a smooth, featureless gate fracture and lower elongation at break per ASTM D638-14 from approximately 80 % to 25 %, even when the pellet moisture is below 0.02 wt%. The cold slug is a separate defect: polymer in the nozzle tip cools below the glass transition temperature during mold-open delay and is injected as a solid plug. The first-shot autopsy must therefore record mold surface temperature, nozzle tip temperature, and screw decompression setting before assigning the failure to wet resin.

Thermal-oxidative versus hydrolytic chain scission in brittle first-shot parts

Polycarbonate degrades by both hydrolysis and thermal-oxidative chain scission, and the corrective action depends on which mechanism dominates. Hydrolytic chain scission consumes water and releases carbon dioxide; it lowers molecular weight without a large initial color shift, so yellowness index per ASTM E313 may remain below 1.5 for short barrel residence. Thermal-oxidative degradation at hot-runner dead zones or overheated barrel surfaces produces yellow-brown discoloration, carbonyl formation, and sometimes crosslinked gel. Melt flow rate measured per ISO 1133-1:2022 cannot distinguish the mechanisms because both increase flow. Notched Izod impact per ISO 180:2019 provides a stronger fracture-based diagnosis: a dried reference may fall from 70 kJ/m² to below 30 kJ/m² in a moisture-degraded first shot, while discoloration remains low. If yellowness index exceeds 2.0, thermal oxidation rather than drying should be the primary inspection target.

Process control for moisture-sensitive polycarbonate should treat the first shot after start-up as an in-process moisture sensor rather than automatic scrap. The dryer is validated by the time series of dew point, return-air temperature, hopper discharge temperature, and feed-throat Karl Fischer results, not by the dryer display temperature alone. If plant relative humidity exceeds 60 %, the transfer system should use dry air at -40 °C dew point, and feed-throat cooling water should be held above the ambient dew point to avoid condensation on the screw shank. Hopper level controls should be interlocked to prevent molding until the residence-height marker is satisfied. Where screw rotation or back pressure is raised to compensate for fill pressure variation, the action can mask but not remove hydrolytic damage; the first shot continues to report the true moisture condition of the feed system.