Polyethylene injection moulding: shrinkage, cycle time and grade selection are the central variables governing part quality and production cost in this semicrystalline material family. Mould shrinkage is defined as the difference between cold cavity dimension and part dimension after conditioning, and is reported under ISO 294-4:2018 on 60 mm × 60 mm × 2 mm plaques or under ASTM D955-21 on 12.7 mm × 127 mm × 3.2 mm bars. Unfilled PE grades generally exhibit mould shrinkage from 0.010 mm/mm to 0.050 mm/mm depending on density, melt flow rate, wall thickness, packing pressure, gate location and mould temperature. Cycle time is governed by solidification of the melt, gate freeze and ejection stability, while grade selection is resolved through melt flow rate under ISO 1133-1:2022, density under ISO 1183-1:2019, tensile properties under ISO 527-2:2012 or ASTM D638-14, and impact resistance under ISO 179-1:2023. No single grade family satisfies all requirements; consequently, shrinkage, cycle time and grade selection are treated as one coupled design space.

Shrinkage magnitude follows semicrystalline solidification and mould restraint

Shrinkage in PE is fundamentally tied to the density increase that accompanies crystallisation. The crystalline domains of polyethylene have a density of approximately 0.997–1.000 g/cm³, while the amorphous fraction is approximately 0.853–0.855 g/cm³ at ambient temperature. As the melt cools from the set barrel temperature of 180–260°C to the mould temperature of 10–40°C, the volume contraction is partially offset by packing. Final mould shrinkage is therefore not simply the density difference; it is strongly sensitive to packing pressure decay after gate freeze. High-density polyethylene with an ISO density of 0.940–0.970 g/cm³ commonly shows mould shrinkage from 0.015 to 0.040 mm/mm. Low-density and linear low-density grades with density 0.915–0.940 g/cm³ may show 0.010–0.050 mm/mm; the wider range for lower-density grades reflects the greater influence of comonomer branching and molecular orientation.

Orientation creates anisotropy in semicrystalline injection mouldings. Melt stretched during cavity filling orients polymer chains, and subsequent crystallisation can lock orientation into the skin, while the core relaxes more slowly. For thin-wall PE parts below 2.0 mm, shrinkage measured parallel to melt flow frequently exceeds transverse shrinkage by 0.002–0.010 mm/mm. In thicker sections, slower cooling permits greater relaxation, and the difference between parallel and transverse shrinkage narrows. Mould restraint is equally important: surfaces in contact with core or cavity cannot shrink freely until ejection, so apparent shrinkage measured after 48 h may increase slightly relative to the value measured at 1 h because post-ejection crystallisation continues. If packing pressure in the cavity is reduced from 80 MPa to 40 MPa, shrinkage can increase, but the exact response depends on gate seal time and part geometry.

Cycle time in PE injection moulding is limited by conductive heat transfer from the melt to the coolant and by gate freeze under packing. For a planar section, cooling time scales with the square of wall thickness. The standard approximation is t_cool = (s² / π²α) ln[(8/π²)(T_melt − T_mould)/(T_eject − T_mould)], where s is wall thickness, α is thermal diffusivity, and the temperatures are melt, mould and ejection values. Using a thermal diffusivity near 0.14–0.24 mm²/s for HDPE at processing temperatures, a 2.0 mm wall moulded at 220°C with a 30°C mould and 90°C ejection temperature is estimated to require 8–14 s of cooling. A 4.0 mm wall may require 25–50 s unless conformal cooling channels or higher mould temperatures are used. These values are starting estimates; on production tooling, actual part extraction is often set by gate seal time, especially with small gates below 0.8 mm diameter, because premature gate freeze stops packing and stabilises the part earlier than the bulk cooling curve would suggest.

Cooling-channel design determines whether the theoretical cooling time is achieved. In hardened tool steel, water lines of 8–12 mm diameter placed at 1.5–2.0 times the wall thickness from the cavity surface are standard. The coolant flow is held in turbulent conditions with a Reynolds number above 10,000; laminar flow reduces heat-transfer coefficient and can extend cooling time by 20–40%. Hot spots near bosses or ribs may require beryllium-copper inserts or high-conductivity tool steels, but published data for PE in these specific configurations is limited and tooling trials are generally required.

What differentiates HDPE, LDPE and LLDPE when balancing flow and toughness?

Grade selection starts with the melt flow rate measured under ISO 1133-1:2022 at 190°C and 2.16 kg. HDPE grades with MFR 4–12 g/10 min are used for thick-wall industrial containers where toughness and environmental stress crack resistance are required; high-flow HDPE grades with MFR 20–30 g/10 min fill thin-wall caps and closures but reduce notched impact strength. LDPE grades with MFR 20–80 g/10 min provide easy flow for complex geometries but yield lower tensile modulus. LLDPE injection grades combine stiffness with higher impact and stress crack resistance, but their narrower solidification range can increase warpage if the mould temperature is not maintained within ±5°C of the set value.

Stiffness follows density. HDPE injection grades with density near 0.955 g/cm³ may exhibit flexural modulus near 1,000–1,400 MPa by ISO 178:2019; LDPE with density near 0.922 g/cm³ may fall near 150–300 MPa. Notched impact strength by ISO 179-1:2023 at 23°C can range from 4–15 kJ/m² for HDPE to 10–50 kJ/m² for selected LLDPE injection grades. Selection therefore cannot be based on flow alone. A cap or closure requiring low-temperature impact and ESCR will usually favour an LLDPE or a bimodal HDPE grade, while a utility part with generous wall section can use a lower-density LDPE grade if the cooling time penalty is acceptable.

When wall thickness exceeds 3 mm and packing cannot compensate for volume change

In sections above 3.0 mm, the thermal gradient through the part creates a frozen skin while the core remains molten. Packing can only compensate for volume change while the gate remains liquid. If the gate freezes before the core, continuing crystallisation generates internal voids or sink marks. This condition is visible in HDPE bosses with a wall thickness to nominal wall ratio above 0.8; ribs should be limited to 50–60% of the adjoining wall thickness to minimise sink. The corrective action is not always longer pack time, but rather a gate diameter large enough to remain open during pack. For PE, gate diameters below 0.8 mm tend to freeze before adequate mass compensation is delivered to thick sections.

Cycle time in thick sections can also be controlled by ejection temperature. PE has a low glass transition temperature, often below -70°C, so parts are not brittle at elevated demoulding temperatures. Ejection at 80–95°C is possible if the part has sufficient rigidity; however, the part continues to shrink after ejection at a rate that depends on ambient temperature and packing orientation. Parts ejected too hot may exhibit dimensional drift of 0.002–0.008 mm/mm over the first 24 h. This reinforces the use of ISO 294-4:2018 conditioning at 23°C ± 2°C and 50% ± 10% relative humidity before accepting a process.

Fast-cycle operations often raise mould temperature to improve skin finish, but this reduces cooling rate. A PE mould set at 40°C instead of 20°C may increase cooling time by 10–20% and shift shrinkage upward because crystallisation continues at lower undercooling. Conversely, very low mould temperatures below 10°C can quench the skin too quickly and produce flow marks or internal stress. Published data for exact shrinkage shift with mould temperature is grade-dependent; process validation should be performed on the production tool.

Processing parameter shifts for fast-cycle thin-wall articles

Thin-wall packaging closures produced from high-flow LDPE or HDPE typically require melt temperatures at the upper end of the grade range to reduce viscosity and an injection velocity sufficient to fill before the flow front solidifies. Screw speeds above 150 min^-1 may generate shear heating that raises actual melt temperature beyond the barrel set value, particularly with high-flow LDPE. The resulting viscosity reduction can shorten fill time but may destabilise the cushion if the non-return valve leaks. Shot-to-shot variance in HDPE shrinkage on a 1,200 kN toggle clamp machine can remain near 0.001–0.003 mm/mm when hydraulic pressure and cushion are held constant, but increases to 0.005 mm/mm or more if non-return valve leakage causes inconsistent pack.

For fast-cycle thin-wall moulds, cooling time may be less critical than robot delay and ejection dwell. A 1.0–1.5 mm wall HDPE cap can cool sufficiently in 5–8 s when the mould is run at 20–30°C, but part handling may require additional delay if the part is too flexible at ejection. Low-density and linear low-density grades with high MFR can reduce fill pressure but often increase mould shrinkage and post-ejection dimensional change. The injection moulder therefore matches grade to both the cavity geometry and the downstream automation, not to melt flow rate alone.