High-density polyethylene (HDPE) grades with solid density between 0.940 g/cm³ and 0.970 g/cm³ undergo mold shrinkage because the melt-to-solid transition is characterized by a pronounced reduction in specific volume. The technical question "Why Does HDPE Shrink During Injection Molding?" is answered primarily by semicrystalline solidification and thermal contraction, not by a single machine setting. Linear mold shrinkage measured on ISO 294-4 or ASTM D955 specimens typically falls between 0.015 mm/mm and 0.040 mm/mm for unfilled HDPE, with thicker sections and constrained flow paths occasionally exceeding 0.050 mm/mm. At a melt temperature near 200 °C, the melt density of HDPE is approximately 0.76 g/cm³ to 0.78 g/cm³, while the room-temperature solid density reaches 0.95 g/cm³ to 0.97 g/cm³; the resulting unconstrained volumetric contraction can exceed 20%. Mold shrinkage is lower than this volumetric value because the packing stage forces additional melt into the cavity and the frozen skin mechanically constrains the core, but the residual linear contraction remains large enough to require tool-steel allowance and process control.

On a production injection molding machine with clamping force of 1500 kN to 4500 kN and a general-purpose polyolefin screw with compression ratio 2.5:1 to 3.5:1, the observed mold shrinkage variation is often larger than the material datasheet value because check-ring leakage, gate freeze timing, and cavity pressure decay differ from laboratory conditions. A worn non-return valve can allow 2% to 5% of the shot volume to bleed back during holding, reducing cavity pressure and increasing shrinkage. Dimensional qualification protocols therefore monitor cushion position, screw recovery time, and cavity pressure repeatability.

Crystallization of HDPE in a cooling mold begins when the quiescent crystallization temperature is reached, normally 115 °C to 125 °C. Differential scanning calorimetry under ISO 11357-3 or ASTM D3418 typically measures crystallinity between 60% and 80%, depending on cooling rate, molecular weight distribution, comonomer content, and nucleating additives. The orthorhombic crystal phase has a theoretical crystal density near 1.00 g/cm³, while the amorphous phase has a density near 0.855 g/cm³; this difference is the fundamental origin of crystallization shrinkage. Once the part cools below the crystallization plateau, solid-state thermal contraction proceeds at a coefficient of linear thermal expansion of 100 × 10⁻⁶ K⁻¹ to 200 × 10⁻⁶ K⁻¹ when measured by thermomechanical analysis according to ISO 11359-2. The combination of crystallization shrinkage and solid-state CLTE creates a processing condition in which dimensional accuracy depends on volumetric compensation before the gate freezes.

What Governs the PVT Response of HDPE During Cooling?

The pressure-volume-temperature path of HDPE is nonlinear across the crystallization plateau. At atmospheric pressure, the specific volume declines gradually in the melt, drops sharply between 125 °C and 110 °C as spherulites form, and then continues to decline in the solid state at a lower rate. Hydrostatic pressure during the holding phase shifts the crystallization temperature upward and compresses the melt, thereby reducing the magnitude of the sharp specific-volume drop. A holding pressure increase from 40 MPa to 80 MPa can reduce linear mold shrinkage by 0.005 mm/mm to 0.015 mm/mm relative to an underpacked condition, provided the gate remains open. The relationship can be expressed as volumetric contraction S_v = (v_melt − v_solid)/v_melt, where v_melt is specific volume at the end of packing and v_solid is specific volume at ambient after demolding. Because the gate must remain molten for pressure transmission, cavity pressure sensors located near the gate and at the last-filled region are used to detect when the pressure curve falls away from the set holding profile; this point is treated as the practical gate seal time.

Polyethylene grade selection changes the solidification path. High-density grades with melt flow rate of 0.2 g/10 min to 50 g/10 min at 190 °C and 2.16 kg under ISO 1133-1 or ASTM D1238 differ in chain length and crystallization kinetics. A low-MFR grade with higher viscosity may develop greater orientation but also requires higher holding pressure to transmit compensation. A high-MFR grade fills thin walls more easily, but the gate may seal earlier in small gates, shortening the volume compensation window. Published data for exact shrinkage values across all commercial grades is limited; material supplier datasheets usually report linear shrinkage after 24 h on ISO 294-4 test plaques, but the reported value is sensitive to wall thickness and gate geometry.

Process settings influence HDPE shrinkage through their effects on crystallinity, pressure decay, and gate seal. The table below compiles representative effects observed in production-scale injection molding of unfilled HDPE using hydraulic or electric injection molding machines with clamping force adequate to maintain the cavity pressure test range.

Among these variables, holding pressure and gate seal time dominate because they directly control mass compensation. On production equipment with cavity pressure sensors, the switch-over from injection to holding is typically set at 95% to 98% of cavity filling, and the holding pressure is maintained until the gate-seal pressure decay is complete. When cavity pressure is not available, screw position and time-based holding are used, but shot-to-shot variation in gate seal and check-ring leakage can produce dimensional scatter beyond 0.002 mm/mm.

Gate Seal, Holding Time, and Cavity Pressure Transmission

Gate seal time for HDPE is determined by the rate of heat removal through the gate cross-section and the latent heat of crystallization. HDPE thermal conductivity is between 0.40 W/m·K and 0.50 W/m·K; this relatively low value slows heat transfer in thick sections. A circular gate with diameter 1.0 mm may freeze in less than 5 s when the mold is held at 15 °C, while a tab gate feeding a 4 mm wall may remain open beyond 10 s. The holding time should therefore be confirmed experimentally by cavity pressure measurement or by incrementally increasing holding time and measuring part mass until mass stabilizes. Once mass stabilizes, additional holding time does not change shrinkage and only extends cycle time. In practice, process engineers add a safety margin of 1.0 s to 2.0 s beyond the observed mass-stabilization point to accommodate material lot variation.

Gate location and size also determine whether shrinkage becomes nonuniform. A gate placed at the thickest section can supply melt during the longest cooling period, but it may also create a local hot spot and differential crystallinity. A gate placed on a thin section may freeze before the thicker core is packed, producing sink or void defects. For multicavity tools, runner imbalance exceeding 5% in fill time can produce cavity-to-cavity shrinkage variation above 0.003 mm/mm; flow balancing with geometrically balanced runners or hot-runner manifold adjustments is therefore part of dimensional control.

The directional shrinkage of HDPE is not isotropic because shear-induced chain orientation, frozen skin layers, and through-thickness crystallinity gradients coexist. In a rectangular plaque, the flow-direction shrinkage after 24 h may be 0.015 mm/mm to 0.025 mm/mm, while the transverse shrinkage may be 0.010 mm/mm to 0.020 mm/mm, depending on injection speed and holding pressure. The skin layer formed against a cold mold contains lower crystallinity and higher frozen orientation, whereas the core cooled more slowly can develop higher crystallinity and greater contraction. This through-thickness structure is one reason why thick HDPE parts can show different shrinkage on opposing surfaces. When the thickness ratio exceeds 1.5:1 across a part, nonuniform cooling can produce warpage even if the average shrinkage is within specification. Simulation tools that incorporate crystallization kinetics and PVT data are used to predict the directionally dependent shrinkage tensor, but experimental mold trials remain necessary because orientation and cooling rate are geometry-specific.

When mold temperature and cooling rate are pushed beyond the stable processing window

If mold temperature is reduced below the dew point, water condenses on cavity surfaces. In an ambient environment of 25 °C and 60% RH, the dew point is approximately 16 °C; running the mold at 10 °C therefore requires dehumidification or mold insulation to prevent surface defects. Cold mold operation also accelerates skin freezing and may reduce bulk crystallinity, which can lower average shrinkage but increase internal stress. The resulting part may exhibit post-mold dimensional change if the part is later exposed to elevated service temperatures above 60 °C, because the interrupted crystallization can proceed slowly. Conversely, a mold temperature of 60 °C slows cooling and increases crystallinity, raising linear shrinkage but reducing frozen-in stress. The choice of mold temperature must therefore balance dimensional magnitude against dimensional stability.

Mold temperature uniformity is critical. A temperature difference across the cavity of more than ±5 °C can generate differential crystallinity and differential thermal contraction, producing warpage. Production tools for HDPE often use circulating water lines with turbulent flow at Reynolds number above 10 000 to maintain consistent heat removal, and the mold is instrumented with thermocouples or thermal imaging during tool qualification. Unfilled HDPE solidifies rapidly at the lower end of the mold temperature range; however, when wall thickness exceeds 6 mm, cooling time becomes the dominant cycle-time variable and the economic benefit of a cold mold is reduced.

Additives change the solidification path of HDPE and therefore change shrinkage. Nucleating agents raise the crystallization temperature and increase nucleating density, which can reduce cycle time but may produce a steeper specific-volume drop and higher mold shrinkage. Mineral fillers such as talc or calcium carbonate lower the linear shrinkage because the inorganic phase has a much lower CLTE than the polymer matrix; filled grades may show linear mold shrinkage below 0.010 mm/mm, but the flow-direction and transverse-direction values can diverge because of filler orientation. Glass fiber at 20% to 30% loading can reduce average shrinkage to the range of 0.003 mm/mm to 0.010 mm/mm, but published data for specific formulations is limited and must be obtained from material supplier datasheets or ISO 294-4 measurements. Recycled HDPE streams with variable density and melt flow rate can exhibit batch-to-batch shrinkage variation exceeding 0.004 mm/mm; blending with virgin material and controlling melt flow rate within ±2 g/10 min narrows this variation.

Semicrystalline solidification proceeds through a narrow processing window

The practical processing window for HDPE injection molding is bordered by excessive melt temperature on one side and premature gate freeze on the other. The melt temperature is commonly set between 180 °C and 230 °C, with high-viscosity grades requiring the upper end to reduce injection pressure. If the melt temperature exceeds 260 °C, oxidative degradation can increase the melt flow rate and reduce mechanical properties, making the shrinkage response less predictable. At melt temperatures below 170 °C, unmelted particles or high viscosity can prevent complete cavity filling and produce anisotropic shrinkage from unrelaxed flow orientation. The stable process is therefore not defined by a single temperature but by a coupled set of melt temperature, mold temperature, injection speed, and holding pressure that maintains cavity mass until gate seal.

HDPE is nonhygroscopic; moisture content of virgin pellets is typically below 0.05 wt%. Drying is not normally required for shrinkage control, but surface condensation on pellets stored in humid environments can introduce splay and minor pressure variability. When condensation is suspected, drying for 2 h at 80 °C is sufficient, but prolonged exposure above 60 °C can oxidize the stabilizer package and shift melt viscosity.

Post-Mold Shrinkage, CLTE, and Measurement Controls

Post-mold shrinkage in HDPE can continue for at least 24 h to 48 h after demolding because secondary crystallization and physical aging occur slowly at ambient temperature. This is the reason ISO 294-4 specifies a conditioning period before measurement; ASTM D955 also reports shrinkage after 24 h and 48 h for many materials. If a part is measured immediately after ejection, the reported shrinkage may be 0.002 mm/mm to 0.005 mm/mm lower than the stabilized value. For critical dimensions, parts are conditioned at 23 °C and 50% RH and measured with calibrated digital calipers or coordinate measuring machines. Dimensional inspection before conditioning can lead to false acceptance.

Tooling dimensional allowance for HDPE therefore requires more than a single shrinkage factor. The cavity is typically cut with a linear mold shrinkage allowance of 0.020 mm/mm to 0.030 mm/mm for unfilled grades, then validated by short-shot or prototype molding and adjusted after ISO 294-4 measurements on a plaque or standard tensile bar. For filled grades or parts with wall thickness variation exceeding 3:1, a single isotropic shrinkage factor is insufficient; the tooling must accommodate directionally dependent contraction. Injection molding simulation that applies PVT data, heat transfer coefficients, and measured melt viscosity is used to calculate the local shrinkage tensor, but the simulation result is not accepted for production tooling until dimensional verification under stable process conditions agrees within 0.002 mm/mm.