When a PET Bottle Blows Out at the Base: Tracing the Cause Upstream

Polyethylene terephthalate (PET) used in stretch-blow-moulded bottles is a semi-crystalline polyester produced from purified terephthalic acid and monoethylene glycol, with bottle-grade resin typically characterized by an inherent viscosity of 0.72 dL/g to 0.84 dL/g under ASTM D4603 or a limiting viscosity number under ISO 1628-5. The transformation from pellet to preform to blown container places the container base at the downstream end of three coupled processes: resin synthesis and solid-state polymerization, preform injection moulding, and two-stage stretch blow moulding. When a PET bottle blows out at the base, the visible failure is not necessarily a moulding defect. It is an accumulation of upstream thermal, rheological, and geometric inputs that converge at the gate vestige and the thin base wall.

Base blow-out may appear as a burst during blowing, a stress-whitened gate region, or a delayed rupture under carbonation pressure. Finished-bottle burst testing under ASTM F1140 records the failure pressure but does not isolate the cause. Drop impact testing under ASTM D2463 may show base cracking when the preform gate region is overcooled or degraded. The upstream trace must therefore begin with the resin lot, pass through drying and melt processing, and then follow the preform through reheat and blow-moulding conditions.

Solid-state polymerization raises bottle-grade PET molecular weight after polycondensation. The resulting molecular weight distribution influences melt strength, stretch response, and base burst resistance. A shift in resin lot IV under ASTM D4603 of only 0.02 dL/g can alter the reheat-blow process window because the preform absorbs infrared energy differently and stretches with a different strain-hardening character. Resin certificates of analysis therefore provide the first upstream boundary of a base failure investigation.

Does the Gate Vestige at the Preform Base Create a Weakness Plane?

During preform injection, the polymer enters through a valve gate located at what will become the bottle base centre. The gate freezes after packing and leaves a small raised or recessed vestige. In high-cavitation preform tools of 48 to 72 cavities, gate-to-gate variation creates differences in local preform base thickness. A gate vestige with sub-millimetre height variation is sufficient to change how the stretch rod contacts the preform base and how the preform distributes into the base mould. Preform wall-thickness profiles are normally checked with an ultrasonic wall-thickness gauge against the tool drawing, but the gate zone may be excluded from routine checks. Consequently, a preform can pass weight and wall-thickness limits and still carry a base asymmetry that fails during blowing.

The gate region is also a point of residual stress. If the gate insert runs hot, the base remains amorphous and can stretch excessively; if the gate insert runs cold, freeze-off is accelerated and the gate may retain a sharp, brittle boundary. Neither condition is visible to the naked eye. Cross-polarized light microscopy and differential scanning calorimetry under ISO 11357-3 can reveal whether the gate region contains unintended crystallinity or thermal history differences. The base failure is therefore not caused solely by blow pressure; it is prepared upstream by the gate thermal balance and by cavity-to-cavity differences in the hot-runner valve-gate system.

Injection moulding parameters further load the base with orientation and enthalpy non-uniformity. Melt temperature during preform moulding is typically maintained between 270 °C and 290 °C. Holding pressure, holding time, and gate cooling time determine how much polymer remains packed into the base before solidification. If holding pressure is reduced to shorten cycle time, base shrinkage may create a local thickness dip. If gate cooling is excessive, the base region may retain high residual stress because the outer skin freezes while the core remains molten. The resulting preform can appear acceptable in total weight but performs differently in the reheat oven. Production-scale experience indicates that cavity-to-cavity variation in gate cooling water temperature on high-cavitation tools can shift the blow-moulding window for individual preforms without triggering a process alarm. Published data for a universal gate-temperature correction is limited because the acceptable range depends on preform weight, gate design, and resin grade.

When Preform Reheat Temperature Gradients Override the Intended Stretch Ratios

A two-stage stretch-blow-moulding line reheats preforms in a quartz infrared oven with pyrometer feedback. The preform surface temperature band for bottle-grade PET is generally 95 °C to 115 °C. If the base is heated above this window, the material can stretch too easily and thin excessively before final mould contact. If the base is heated below this window, the stretch rod may load the gate zone without sufficient polymer mobility, producing stress whitening or a punch-through at the gate. The base blow-out is thus related to the preform temperature gradient, not only to the final blow pressure.

Typical bottle designs apply an axial stretch ratio of 2.2 to 3.0 and a hoop stretch ratio of 3.5 to 4.5. At the base, the effective hoop stretch is lower than the sidewall because the preform geometry changes near the gate. Small temperature deviations in the base region produce disproportionately large thickness changes. A reheat oven with closed-loop control may hold the surface temperature to a narrow band, but start-up after a colour change, label-free preforms, or a change in preform storage temperature can create a base-specific temperature error that is not detected by the sidewall profile. When the base then fails under ASTM F1140, the corrective action may require changing the oven lamp settings or the preform conditioning time upstream rather than altering blow-moulding pressure alone.

The pre-blow pressure sequence is another upstream blow-moulding parameter with direct base consequences. If pre-blow pressure is introduced too early or at too high a level, the preform expands before the stretch rod completes its axial travel. The material is then pushed into the base mould before orientation is established, producing an off-centre gate or a thin base wall. If pre-blow pressure is too late or too low, the preform remains thick near the gate, and the base may not replicate the petaloid or champagne geometry. Process engineers trace these failures by mapping base wall thickness against pre-blow timing, pre-blow pressure, and stretch-rod velocity. Published data for a universal pre-blow pressure is limited because the correct value depends on the preform weight, reheat profile, and base mould design. The upstream cause is therefore not a single pressure setting but the interaction between the preform temperature profile and the blow-moulding sequence.

Drying, Melt Temperature, and Intrinsic Viscosity Retention Limits

Moisture entering the injection barrel is one of the most direct upstream causes of a brittle base. PET is hygroscopic; amorphous pellets stored at 50% relative humidity can pick up enough surface moisture to degrade the melt. Hydrolysis cleaves ester linkages and lowers molecular weight. Bottle-grade PET should be dried to a moisture content below 50 ppm (0.005 wt%) as measured by ASTM D6869 or ISO 15512. Desiccant dryers with a dew point of -40 °C or lower, an air temperature of 160 °C to 180 °C, and a residence time of 4 h to 6 h are typical production-scale settings. If moisture remains above 100 ppm, the inherent viscosity can fall below 0.65 dL/g, and the base may fail by brittle cracking instead of ductile yielding.

Melt temperature also controls inherent viscosity retention. Preform injection at 270 °C to 290 °C keeps thermal degradation within a controllable range. Above 300 °C, acetaldehyde generation and IV loss accelerate sharply. Acetaldehyde is monitored by ASTM F2013; water-grade preforms often require residual acetaldehyde below 8 ppm. Although acetaldehyde is primarily a sensory specification, elevated acetaldehyde indicates thermal damage to the polymer backbone. That thermal damage appears later as a base with reduced burst resistance because the load-bearing ester linkages have been shortened. The same upstream thermal history can produce a preform that fills easily but cannot survive the orientation stresses of the blow process.

Preform storage between injection and blow moulding is a frequently overlooked upstream variable. Preforms that are not moisture-protected can regain enough water at the surface to alter the reheat response and generate additional hydrolysis during the brief stretch-blow process. Although the exposure time is shorter than pellet drying, the surface area of a cold preform is high, and the base gate region can absorb moisture at a different rate than the sidewall because of local crystallinity differences. A preform storage history that differs between reference and production lots can therefore produce intermittent base failures even when the resin certificate of analysis remains unchanged.

A Different Upstream Input: Copolymer Content and Crystallization Behavior

Bottle-grade PET is often copolymerized with 1–2 wt% isophthalic acid or cyclohexanedimethanol to slow crystallization and widen the processing window. A resin with insufficient comonomer may crystallize prematurely in the preform base during injection or reheat, producing a hazy base with lower stretchability. A resin with excessive comonomer may lower the glass transition and soften the base to a level that reduces burst pressure. The base-forming response is therefore set long before the preform is blown, in the resin formulation and solid-state polymerization step. Differential scanning calorimetry under ISO 11357-3 is used to confirm that the amorphous preform base does not contain a significant melting peak before blow moulding. If a melting peak appears, the preform base has been overwritten by crystallinity, and the blow-moulding machine must work against a heterogeneous structure.

Solid-state polymerization further defines the molecular weight distribution and carboxyl end-group concentration. The base wall that survives ASTM F1140 burst testing is a function of chain length, orientation, and thickness. When the resin supplier changes the comonomer level or the SSP residence time without this being communicated, the blow moulder may observe a sudden rise in base blow-outs even though the preform weight and visual clarity remain unchanged. The trace upstream therefore includes change-control across the resin lot and verification against the original preform qualification data.

On a production-scale two-stage line, intermittent base blow-outs after a preform lot change require a lot-specific trace. The upstream analysis should compare the failing preform lot with the reference lot using ASTM D4603 for inherent viscosity, ASTM D6869 or ISO 15512 for moisture, ASTM F2013 for acetaldehyde, ISO 11357-3 for thermal history, and ASTM F1140 for finished-bottle burst resistance. The comparison is not sufficient if it stops at the resin certificate of analysis; preform gate geometry, base wall thickness, and reheat oven settings must also be inspected. Published data for universal acceptance limits across bottle designs is limited because each container design has a different base thickness and pressure requirement. The base blow-out therefore functions as a downstream signal that upstream conditions have moved outside the qualified processing corridor.