Why Can Changes in HDPE Resin Increase Production Scrap?

Across a multi-machine plant running high-density polyethylene, scrap is frequently attributed to machine drift or operator error; however, the dominant root cause may be a change in the HDPE resin supply. Resin lot-to-lot variation can shift viscous flow, melt strength, crystallisation kinetics, additive chemistry, pellet conveying and regrind composition without violating the broad datasheet ranges. The result is not a single scrap signature but a cluster of failures—wall-thickness variation, parison curl, sheet edge curl, gate blush, black specks, short shots and dimensional drift—that appears only after the new lot is already in the hopper. The mechanisms below isolate those process changes and the corresponding test methods.

Molecular Weight Shifts and Melt Flow Ratio Instability

An HDPE blow-moulding grade specified at 0.30 g/10 min under 2.16 kg at 190 °C may arrive at 0.22 g/10 min or 0.36 g/10 min and still remain inside a commercial specification. The melt flow rate measured per ISO 1133-1 is only a single point on the shear-viscosity curve; a change in average molecular weight or molecular weight distribution can produce a much larger shift in die-land flow. On a single-screw extruder with 30:1 L/D and a grooved feed section, a fall from 0.30 g/10 min to 0.22 g/10 min raises head pressure, reduces throughput at constant screw speed, and increases residence time. The longer thermal history generates gel particles and discolouration, and thin-walled bottle scrap rises even though the resin remains within specification.

Melt flow ratio, the quotient of 21.6 kg and 2.16 kg MFR values per ISO 1133-1, is a practical indicator of molecular weight distribution. A broadened distribution increases elastic recovery and die swell, while a narrowed distribution reduces melt strength and parison stability. In extrusion blow moulding, the narrowed distribution may induce parison sag and pinch-off seam holes; the broadened distribution may cause uneven wall thinning near the parting line and tail scrap. Both conditions can occur without any change in die gap, core pin temperature, or clamp force.

Capillary rheometry per ASTM D3835 should be used to compare shear viscosity at 100 s⁻¹ and 1,000 s⁻¹ across suspect lots. A lot with lower high-shear viscosity may reduce melt pressure but exhibit sharkskin at lower screw speeds; a lot with higher low-shear viscosity may trigger head-pressure alarms and require barrel-temperature changes outside the validated window. For high-molecular-weight HDPE blow-moulding grades, the practical melt-temperature window is often narrow—on the order of 10 °C to 15 °C—before oxidation or melt fracture appears. Published data for this specific configuration is limited, but the direction of change is consistently observed on accumulator-head machines with production rates near 300 kg/h.

In sheet extrusion of HDPE for thermoformed trays, an incoming lot with density 0.958 g/cm³ replacing a nominal 0.952 g/cm³ can generate edge curl and roll-stack fouling even when the supplier’s density specification permits the change. Density measured by ISO 1183-1 or ASTM D1505 is not a melt-flow parameter, but it correlates with crystallinity and heat release during cooling. The higher-density material solidifies faster at the roll stack and develops different thermal shrinkage in the machine and transverse directions. Scrap arises when trim knives cannot track the curled sheet or when formed trays exceed stack-fit tolerances after 24 h of ambient storage.

Thermoforming tools using plug assistance are sensitive to quiescent crystallisation rate. A density shift of 0.004 g/cm³ may be sufficient to change the surface temperature at plug contact because the crystalline skin cools and solidifies earlier. The result is uneven wall distribution in corner radii and rejected parts from minimum-wall thinning. Differential scanning calorimetry per ISO 11357-3 can quantify the crystallisation peak shift; if the peak moves by 3–5 °C, the cooling-water flow and mould temperature settings may require re-validation before the next lot is accepted.

In injection moulding, density changes alter mould shrinkage. HDPE linear mould shrinkage commonly ranges from 1.5 % to 4.0 % depending on wall thickness, gate geometry and holding pressure. A higher-density lot can produce parts at the upper end of that range and create dimensional scrap in closures with threaded sealing diameters. Shrinkage measured per ASTM D955 or ISO 294-4 must be repeated on the actual gate geometry, because shear-induced orientation and post-mould crystallisation differ between edge-gated and hot-runner tools.

What Role Does Comonomer Type and Short-Chain Branch Distribution Play in Bottle and Film Scrap?

High-density polyethylene is not a single chemistry. Changes among butene, hexene and octene comonomers alter short-chain branching length and distribution while density and melt index may remain unchanged. A switch from a butene-based HDPE to a hexene-based HDPE in a blown film line can alter strain-hardening behaviour, bubble stability and dart impact before off-line testing is complete. The scrap mechanism is usually film tear at the frost line or oscillation in bubble diameter that moves beyond the automatic air-ring correction range.

Blow-moulding grades with similar 0.950 g/cm³ density and 0.30 g/10 min melt flow rate can differ in environmental stress crack resistance because comonomer placement controls tie-molecule density. A lot that reduces short-chain branching in the high-molecular-weight fraction may extrude acceptably but produce bottles that fail top-load after filling or stress-crack at the neck. ASTM D1693 testing of the resin gives a comparative ranking, but the relationship between the laboratory notched specimen and a moulded bottle thread is geometry-dependent and temperature-dependent.

Catalyst residue changes are another hidden variable. Chromium-catalysed HDPE often has a broad molecular weight distribution and higher die swell than a Ziegler-Natta or metallocene grade at similar melt index. If a processor qualifies a chromium-catalysed film resin and then receives a metallocene resin with the same nominal MFR, the extruder may run hotter, the bubble may sag, and the dart impact may improve while the coefficient of friction shifts. Without resin identification and melt-rheology fingerprinting, the line may produce film that meets tensile specifications but fails downstream bag conversion due to slip, tear or heat-seal temperature drift. Kinetic coefficient of friction can be monitored per ASTM D1894.

In high-speed cap and closure injection moulding, a change from one HDPE supplier lot to another may involve a different acid-scavenger package, a lower phenolic-to-phosphite ratio, or a shift in slip additive from erucamide to oleamide. These differences do not alter the 0.952 g/cm³ density or 20 g/10 min melt flow rate, but they can change melt lubrication, screw recovery time, and gas evolution. Scrap appears as nozzle plate-out, mould deposits, gate blush or colour delta beyond ΔE 1.5 on a benchtop spectrophotometer.

Oxidative induction time measured by ISO 11357-6 should be part of incoming resin release for any HDPE lot that will be held in a hot runner above 220 °C for more than 20 min. A reduction from 45 min to 12 min at 200 °C may remain within some datasheets, yet the hot-runner tip can generate carbonised material that breaks free intermittently and contaminates subsequent shots. The resulting black-spec scrap is often incorrectly blamed on thermocouple drift or back-pressure settings. In colour-critical HDPE, combinations of hindered amine light stabilisers and phenolic antioxidants can also produce yellowing at elevated processing temperatures; this is an additive incompatibility rather than a melt-flow failure.

Surface-active additives such as erucamide and antistatic agents migrate to the part surface over time. When a resin change increases slip concentration, printing and adhesion operations may fail after 24–48 h of storage; when it decreases slip concentration, caps may not unscrew at the specified removal torque and the line may resort to overdosing external lubricant. Both directions produce scrap in secondary operations that is visible only after the primary moulding line has released the lot.

When Pellet Fines and Bulk Density Shift on a Single-Screw Feed Throat

Pellet geometry is often omitted from resin change assessments. HDPE is received as cylinders, spheres, lenticular pellets or irregular granules; each shape has a different angle of repose, bulk density and feed-throat behaviour. A change from spherical pellets with bulk density 0.58 g/cm³ to irregular granules with bulk density 0.53 g/cm³ may not affect melt-phase rheology but can destabilise the feed section of a smooth-bore extruder with 28:1 L/D. The screw flights may fill partially, the melt temperature profile may oscillate, and gauge variation becomes non-uniform.

Fines generation during pneumatic conveying is a frequent root cause. HDPE pellet cutting can produce fines and streamers after railcar unloading, silo level changes, or dense-phase transfer. Fines above 500 ppm by weight can bridge in the feed throat, starve the screw, and produce shot-mass drift in injection moulding or thickness bands in sheet. Bulk density measured per ASTM D1895 or pourability per ISO 6186 should be recorded for each silo sample; a deviation greater than 5 % from the qualified value may require conveyor speed or feed-hopper bridge-breaker changes.

In injection moulding machines with volumetric dosing and screw recovery, bulk density variation translates directly into shot size variation because screw-back position measures volume, not mass. A lot with lower bulk density can produce short shots and flashing in alternating cavities; the cavity with a hot spot may flash while another cavity shrinks. This scrap is often misattributed to clamp force or check-ring wear. The correct response is mass-based shot control or gravimetric hopper feeding, not repeated process recalculations.

Although HDPE is not hygroscopic, condensation on cold pellets stored outdoors at humidity above 60 % RH can introduce surface water that causes splay and voids. Pre-drying at 80 °C for 2 h in a desiccant hopper is sufficient for surface moisture removal and is a useful diagnostic when splay appears only after silo temperature swings.

In plants blending post-consumer recyclate into HDPE at 25 wt% to 50 wt%, the incoming recyclate is itself a resin change that cannot be controlled by supplier datasheets. Polypropylene contamination, EVOH barrier fragments, label adhesive, and mixed high-density grades create local viscosity domains, gas evolution, and gels. A virgin HDPE lot that was stable at 190 °C to 210 °C can produce splay, black specks and hot-runner clogging when the blend contains thermally degraded recyclate with an OIT below 5 min.

The melt filtration pressure trend is an early indicator. On a single-screw extruder fitted with a 200 mesh screen pack and screen changer, a pressure rise above 35 bar over 4 h may indicate gel or contamination carryover from a changed recyclate source. Force-frequency data from the screen changer can separate gradual filter plugging from sudden rupture of filter media. Scrap in this regime is not random; it is correlated with the elapsed time since screen indexing.

In recyclate blends, density readings are ambiguous. A drop from 0.956 g/cm³ to 0.948 g/cm³ may indicate polypropylene contamination, while a rise above 0.965 g/cm³ may indicate mineral-filled or coloured flake. Both shifts alter crystallisation and shrink in injection-moulded crates or pallets. Differential scanning calorimetry per ISO 11357-3 can identify a secondary melting peak near 160 °C from polypropylene contamination even when the overall density remains within the HDPE specification.

Extrusion coating lines running HDPE onto paper or board respond to narrow changes in draw resonance and neck-in. A resin lot with lower melt tension, even at identical 7 g/10 min melt flow rate and 0.956 g/cm³ density, can widen neck-in and shift edge-bead formation. Scrap generation is immediate: the coated web may exceed the substrate trim width, causing edge splice roll contamination. Capillary rheometry per ASTM D3835 and melt tension measurement are used to compare lots before high-speed coating starts.

The curtain stability between die exit and nip is governed by extensional rheology rather than shear viscosity alone. A change in long-chain branching or a high-molecular-weight tail can convert a stable curtain into a sagging sheet that folds at the nip. On a coating line running 300 m/min, the onset of draw resonance may appear as sinusoidal gauge bands that are detectable only after the roll is slit. The scrap is not off-line pellets; it is out-of-spec roll stock that must be rewound or scrapped.

Crystallisation Rates Shift Shot-to-Shot Mass in Cap Moulding

Injection moulding scrap from HDPE is often assumed to be a temperature-control problem, but a resin change that replaces a lot with different nucleation density can shift crystallisation kinetics without a shift in melt flow rate. Fast crystallisation reduces gate seal time, changes the pressure window for packing, and increases the sensitivity of part mass to holding pressure. A cap tool with 64 cavities may begin to show mass sigma above 0.020 g after a resin change, triggering dimensional scrap at thread closure.

Nucleating agents, talc residues, or catalyst carryover can accelerate crystallisation. The crystallisation peak temperature measured by ISO 11357-3 may move from 116 °C to 121 °C. That 5 °C shift is small in absolute terms but changes the amount of supercooling available during mould filling and the time available for polymer chains to relax before solidification. Parts may show higher frozen-in orientation, greater shrinkage anisotropy, and warpage on ejection.

Process engineers can compensate by raising mould temperature from 15 °C to 25 °C or by reducing holding pressure, but both changes move the process away from the qualified window. Scrap from dimensional drift is best isolated by measuring part mass, cavity-to-cavity mass spread, and post-mould shrinkage after 48 h. If the new resin lot exhibits a crystallisation peak shift greater than 3 °C compared with the qualification lot, the lot should be segregated or processed under a revised protocol.