The Bend Went Cloudy: A Polycarbonate Thermoforming Failure Walk-Through

Polycarbonate sheet at 3.0 mm nominal gauge was formed on a three-station vacuum forming machine with zoned ceramic heaters. The sheet had been dried in a desiccant dryer at 120 °C for 4 h and formed at a surface temperature of 175 °C against an aluminum female tool held at 85 °C. The formed part was transparent except for a clouded band 18 mm wide along the sharpest bend radius. Haze on the flat section measured 1.1% per ASTM D1003-13, while haze on the bend line measured 13.6%. Local thickness at the defect was 1.4 mm, compared with 2.9 mm in the adjacent flat wall.

Visual inspection under D65 illumination showed no silver streaks on the flat areas. The defect remained after heating the clouded bend to 130 °C for 2 h, indicating that the haze was not simply condensed surface moisture. A scrap sample was taken across the bend line for Karl Fischer titration, melt volume-flow rate measurement, differential scanning calorimetry, and optical haze testing.

Why Does Localized Haze Appear Only Along the Bend Line?

Localized bend haze in amorphous polycarbonate separates into three primary mechanisms. The first is residual moisture vaporization. At a sheet surface temperature of 175 °C, the saturation pressure of water is approximately 0.89 MPa. If the sheet contains more than 0.02% residual moisture, steam bubbles can nucleate inside the softened polymer. Because the bend line experiences the highest tensile strain and the greatest wall thinning, any voids elongate and become optically visible. The second mechanism is stress whitening from microcavitation and craze formation. Polycarbonate develops shear bands and microvoids under tensile true strain above roughly 0.5 to 0.7, depending on molecular weight, temperature, and strain rate. The third mechanism is surface roughening or volatile condensation transferred from the tool surface. In this failure walk-through, residual moisture was 0.021% by Karl Fischer titration at 160 °C, which is marginal. The dominant cause was determined to be tensile microcavitation at the bend line because annealing produced less than 1.0 percentage point reduction in haze.

The distinction between moisture haze and strain haze can be tested by thermal annealing. Moisture voids partially collapse at temperatures above the glass transition and haze decreases. Strain-induced microvoids remain stable because they are separated by oriented polymer chains that resist healing at 130 °C for 2 h. In this case, bend haze measured 13.6% before annealing and 12.9% after annealing, confirming microcavitation rather than residual moisture alone.

Dryer Dew Point, Residual Moisture, and Sheet Thickness

Polycarbonate absorbs moisture from ambient air. At 23 °C and 50% RH, equilibrium moisture content is approximately 0.15–0.20% by mass according to ISO 62:2008. At 23 °C and 80% RH, moisture uptake can exceed 0.30%. Thermoforming and extrusion require residual moisture below 0.02%. This is normally achieved in a dehumidifying desiccant dryer with a dew point of −40 °C to −30 °C at the drying hopper inlet. A rise in dew point to −20 °C reduces the moisture gradient and can leave 0.03–0.05% residual moisture after the same residence time.

Drying kinetics for 3.0 mm sheet follow Fickian diffusion. The time required to reach the core moisture target scales with the square of sheet thickness. A 3.0 mm sheet dried at 120 °C for 4 h from 0.15% initial moisture will normally pass a 0.02% limit. The same sheet stored in a humid warehouse until 0.35% may require 8–10 h at 120 °C. Overdrying above 130 °C for more than 12 h can initiate surface yellowing and reduce impact performance. The drying hopper should be insulated, and the return air cooler should maintain a consistent dew point below −30 °C.

Residual moisture and thermal history can be separated by melt volume-flow rate measurement according to ISO 1133-1:2022 at 300 °C/1.2 kg. A virgin polycarbonate resin with a nominal MVR of 9–10 cm³/10 min may shift upward by 1–3 cm³/10 min if hydrolysis has reduced molecular weight. A shift greater than 10% from the virgin lot is evidence of molecular weight loss. In this case, MVR of the bend line scrap was 10.8 cm³/10 min compared with 9.3 cm³/10 min on the flat edge, a 16% increase. This supports chain scission at the bend, likely due to the combined effects of residual moisture and high local forming stress.

When Surface Temperature Overshoots at Thin Sections

Polycarbonate sheet requires a surface temperature between 165 °C and 190 °C for vacuum forming. Thick sheet often forms at 170–180 °C. At the bend line, the sheet thins faster because the local draw ratio is highest. If the heater control is biased toward the flat zones, the thin bend area can overheat to 200 °C or higher. At these temperatures, residual water reacts rapidly with carbonate linkages. Even 0.02% moisture can generate microvoids when local strain rate is high. Published data for this specific configuration is limited; however, the combination of high local temperature, marginal moisture, and tight bend radius creates a narrow processing window.

Infrared pyrometers or embedded thermocouple probes are used to map sheet surface temperature. A temperature spread across the sheet greater than ±5 °C can cause the thin section to draw inconsistently. Ceramic heaters with zoned quartz lamps are typically set to compensate for edge losses and thickness variation. In this walk-through, the heater zone over the bend was reduced by 7 °C after the failure. The aluminum female tool was held at 85 °C, within the typical 80–120 °C range for polycarbonate thermoforming. Mold temperatures below 70 °C produce early freeze-off and reduce part replication; mold temperatures above 130 °C risk release marks and cycle-time loss.

Vacuum pressure of −0.08 MPa to −0.09 MPa was applied after the plug assist reached 75% of stroke. The bend radius was 4.0 mm with a sheet thickness of 3.0 mm, giving a radius-to-thickness ratio of 1.3:1. This is below the commonly recommended minimum of 2:1 for unassisted polycarbonate forming. A tight bend radius increases local draw ratio and haze risk. Opening the bend radius to 7.5 mm reduced the local draw ratio and increased the radius-to-thickness ratio to 2.5:1.

Corrective action on the line focused on three variables. The dryer dew point was lowered from −25 °C to −40 °C, residence time was extended from 4 h to 6 h, and the bend radius was opened from 4.0 mm to 7.5 mm. The heater zone over the bend was reduced by 7 °C to prevent local overheating. After these changes, bend-line haze measured 2.8% and local thickness increased to 2.3 mm.