Menu
The certification of a high-density polyethylene jerrycan under the United Nations design type test regime subjects the as-moulded article to a sequence of mechanical challenges that bear no direct resemblance to a simple tensile coupon. A jerrycan classified as 3H1 is a plastics jerrican with a non-removable head, and for liquid dangerous goods it must pass the drop test specified in UN 6.1.5.3. The drop test requires three specimens, or one specimen dropped in three orientations: one drop on the bottom, one drop on the side, and one drop on the top or side seam. The pass criterion is straightforward: no leakage may occur. For liquids with relative density not exceeding 1.2, the fixed drop heights are 1.8 m for packing group I, 1.2 m for packing group II, and 0.8 m for packing group III. Parison sag variation enters this qualification path because the blow-moulded wall thickness distribution, particularly around the shoulder, handle, and pinch-off weld, determines whether any given orientation survives the impact without leakage. The relationship between parison sag and drop test performance is not a single curve; it is a convolution of local thickness, molecular orientation, weld-line strength, and high-rate deformation response.
Extrusion blow moulding of HDPE jerrycans begins with an accumulator head that fills during screw recovery and then discharges a hollow parison through an annular die. The parison is extruded downward under gravity, and from the instant it exits the die swell region, axial tension is generated by the weight of the hanging melt. Sag is the time-dependent elongation of the parison before mould closure. In a 20 L jerrican tooling, the parison hang time may be several seconds, and because HDPE melt is viscoelastic, the local reduction in cross-section is not uniform along the parison length. The upper end of the parison remains connected to the die and is drawn into a thinner annulus earlier, while the lower end continues to accumulate melt and can thicken before pinch-off. When the mould closes, the wall thickness of the blow-moulded part is therefore biased: zones formed from the upper parison tend to be thinner around the shoulder and handle, while the lower parison contributes more material to the bottom pinch-off weld. Variations in sag from shot to shot shift this bias, producing intermittent thin regions that may pass a static wall-thickness audit but fail under the drop impact. Published data for a direct quantitative correlation between sag length and UN drop test failure rate in commercial 3H1 jerrycans is limited, but industrial qualification commonly links chronic drop failures at the top or side seam to excessive sag-induced thinning in the upper parison segments.
Shot-to-shot parison sag variation is rarely dominated by a single controlled variable. The accumulator head temperature, die gap, extrusion rate, shot weight, resin lot, and ambient airflow around the parison all contribute. Commercial HDPE blow moulding grades used for UN-rated jerrycans are frequently bimodal high-molecular-weight copolymers with density between 0.945 g/cm³ and 0.955 g/cm³ and melt flow index between 0.2 g/10 min and 0.5 g/10 min at 190 °C/2.16 kg per ISO 1133-1:2022. These resins exhibit high zero-shear viscosity and extensional strain hardening, which resist sag, but they remain sensitive to melt temperature drift. A rise in melt temperature from 190 °C to 210 °C reduces viscosity and accelerates sag even if the programmed die gap is unchanged. At the same time, increasing temperature can improve pinch-off weld healing because the two melt fronts retain molecular mobility for longer during clamp closure. This creates a processing conflict: the same temperature shift that worsens sag-driven thinning may improve weld integrity. Drop test outcomes therefore cannot be optimised by temperature alone, and industrial lines must control melt temperature within tight bands, typically on the order of ±2 °C to ±3 °C, while simultaneously monitoring shot weight and parison length. Accumulator-head machines commonly run with screw length-to-diameter ratios between 24:1 and 30:1, and die gaps for a 20 L jerrican tooling are typically set between 1.5 mm and 3.0 mm. These values may be adjusted during initial qualification, but once a design type approval is issued, changes in die gap or material lot that alter sag behaviour can invalidate the basis of the approval if not revalidated.
For liquid-filled packagings with a relative density greater than 1.2, UN 6.1.5.3.5 replaces the fixed drop height with a density-scaled value. The formula uses the relative density of the intended liquid as a multiplier against the packing group base height, and the result is rounded to the first decimal place. The increase is mechanically significant for HDPE jerrycans because impact energy rises with drop height and the additional hydrostatic pressure during impact increases stress on the sidewall and top closure. A packaging group II jerrican carrying a liquid with relative density 1.45, for example, would be dropped from 1.45 × 1.2 m = 1.74 m, rounded to 1.7 m, instead of 1.2 m. A packing group I jerrican carrying a liquid with relative density 1.9 would be tested from 3.42 m, rounded to 3.4 m. These adjusted heights make thin regions produced by parison sag more critical because the same local thickness must now absorb higher kinetic energy without cracking or tearing.
| Packing group | Base drop height for liquids ≤ 1.2 relative density | Adjusted drop height formula for relative density > 1.2 |
|---|---|---|
| I | 1.8 m | d × 1.8 m |
| II | 1.2 m | d × 1.2 m |
| III | 0.8 m | d × 0.8 m |
The density adjustment is not a minor regulatory detail; it alters the failure ranking of orientations. A jerrican with a marginally acceptable top drop at 1.2 m may fail at 1.7 m if the upper parison was sag-thinned. Similarly, a bottom pinch-off weld that survives the standard side drop may split at higher energy if the weld interface contains localised cold spots caused by parison position variation. Designers of UN-rated HDPE jerrycans must therefore evaluate worst-case contents with the highest specified relative density and must account for the fact that the drop test height is set by the product to be transported, not by the nominal container design alone. During production, lot-specific density verification of the filled chemical is performed separately from the packaging test, but the packaging manufacturer must know the maximum relative density claimed on the approval certificate. If the certificate lists a maximum density of 1.6, the design must satisfy the corresponding adjusted drop height.
Because the drop test converts gravitational potential energy into a short-duration impact strain, failure in an HDPE jerrycan is governed less by tensile yield strength measured at quasi-static rates than by high-strain-rate localisation at thickness discontinuities. HDPE is strongly strain-rate sensitive: yield stress rises as loading rate increases, but the ductile-to-brittle transition and the localisation of strain into thin bands can reduce the energy absorbed before failure. A thin shoulder section formed by sag may not fail during low-speed flexing or hydraulic pressure testing, yet it can crack during the 1.2 m or 1.8 m drop because the impact event suppresses the large-scale yielding that would otherwise distribute deformation. At the pinch-off weld, the issue is compounded by molecular orientation and welding history. When the mould halves compress the parison, two hot melt fronts meet under pressure; if the local temperature has decreased because the lower parison cooled during extended sag, chain interdiffusion across the weld interface is incomplete. The resulting weld may display weak transverse strength even though the surrounding wall thickness is acceptable. In drop orientations where the weld is placed in bending or tension, a brittle crack can initiate at the weld line and propagate along the bottom or side. Instrumented impact data from blow-moulded containers tested to ASTM D2463-15 or ISO 6603-2 are often used to characterise the high-rate response of HDPE plaques and bottles, but UN drop testing remains the governing pass/fail method for dangerous goods packaging. Published data for a specific parison sag-to-drop outcome correlation under UN 6.1.5.3 is limited; most commercial qualification programs therefore rely on design type tests supplemented by destructive wall-thickness mapping and weld-strength audits.
Instrumented drop towers used for UN type qualification combine a quick-release mechanism, a rigid impact surface, and high-speed video acquisition with accelerometer readouts on the drop carriage. The impact surface must be hard, horizontal, flat, and non-resilient. The drop orientation is adjusted by supporting fixtures or by manually orienting the jerrican before release. High-speed video at frame rates of 2000 fps to 10000 fps can capture the deformation sequence: initial contact, elastic flexure, buckling of the sidewall, compression of the top, and rebound. Accelerometer data reveal peak deceleration and the duration of impact, which for a filled jerrican may be on the order of 10 ms to 30 ms. The load cell in the drop carriage, if fitted, measures the dynamic force transmitted through the package. These measurements are not required by the UN pass criterion, but they help identify whether a failure initiates at the first contact region or later during rebound. When a jerrican fails, the crack often initiates at the pinch-off weld on the bottom perimeter or at the sharp transition between the handle and the sidewall. Both regions are sensitive to parison sag because the pinch-off weld forms from the lower parison end, and the handle area is formed from the upper parison that may have thinned while hanging. A failure map can be constructed by superimposing the crack location on a grid of ultrasonic wall-thickness readings taken around the failed container. This map frequently shows that the crack follows a path where the average thickness falls below a design minimum, and the minimum is often co-located with the parison segment that exhibited the greatest sag-related elongation.
The value of instrumented drop testing lies not in replacing the visual leak detection criterion but in separating failures caused by material contamination from failures caused by geometry. A crack that initiates at a black speck or degraded regrind particle may have a different root cause than a crack that initiates at a uniformly thin band. In the first case, the corrective action involves filtration, regrind ratio control, and screw-cleaning procedures; in the second, it involves parison length control, die-gap adjustment, and melt-strength management. On production-scale shuttle blow moulders, the interaction between the mould closing mechanism and the parison is not perfectly repeatable. Mould closure speed can affect the point at which the parison is captured, and small differences in closure timing change the amount of sag before the mould halves contact the parison. A jerrican that passes the drop test with a shot produced at the start of a production run may fail with a shot produced after the accumulator head temperature has drifted upward. Process engineers therefore monitor the parison tip position with laser or linear variable differential transformer sensors at a fixed time after extrusion and compare it with the shot weight recorded by the accumulator. The comparison gives an indirect indication of whether sag variation is being driven by melt temperature or by shot-size drift. If the shot weight is stable but the parison length increases, the likely cause is reduced melt viscosity or increased hang time. If the shot weight changes while the parison length remains constant, the die gap or accumulator fill may be drifting. Both patterns can produce a drop test failure, but the corrective actions differ.
In plants where ambient relative humidity exceeds 60 %, HDPE feedstock can carry surface moisture into the extruder, although the effect on viscoelastic sag is usually smaller than the effect of melt temperature drift. HDPE is generally regarded as non-hygroscopic, and pre-drying is not mandatory for standard blow moulding grades under normal factory conditions. However, surface moisture on pellets can produce small surface streaks or voids in the parison, and these defects may act as stress concentrators during the drop test. If a pelleted batch is exposed to condensation, a desiccant hopper drier set at 80 °C for 2 h is sometimes used, but this is a corrective measure rather than a standard requirement. More significant for sag are resin lot changes within the same commercial grade. Different lots of nominally identical bimodal HDPE can exhibit variations in molecular weight distribution, long-chain branching, and comonomer distribution that alter zero-shear viscosity and extensional strain hardening. The melt flow index may remain within specification, but a lot with a lower high-load melt flow index or a higher storage modulus at low frequency may sag less at the same melt temperature. This is why incoming resin quality control for UN-certified jerrycans often includes capillary rheometry or dynamic oscillatory shear testing according to ISO 6721-10, even when the supplier reports only the nominal melt flow index. A lot that passes the standard ISO 1133-1:2022 check can still produce sag-driven drop failures if its melt strength is at the low end of the grade envelope.
The introduction of laser-based parison length sensors and closed-loop shot-weight control has reduced, but not eliminated, the influence of sag variability on drop test outcomes. In a modern accumulator-head machine, the parison tip position is measured at a preset time after the start of extrusion, and the controller can trim the die gap or accumulator push-out speed for the next shot. A 100-point die gap programmer can also generate a wall-thickness profile along the parison length, compensating for known sag behaviour by opening the die gap at the upper sections and closing it at the lower sections. The programmer output is based on a model of how the parison will sag after extrusion, but the model must be updated when the resin lot, melt temperature, or shot size changes. The sensor itself does not measure local wall thickness after sag; it measures only the position of the parison tip. Similarly, shot-weight control measures the total mass discharged but not how that mass is distributed vertically. A controlled shot weight of 1.7 kg for a 20 L jerrican does not prevent a top-thin condition if excessive sag redistributes the mass downward before the mould closes. Closed-loop sag compensation can therefore reduce low-frequency drift and improve process capability, but it cannot convert a resin with poor melt strength into a drop-test-proof jerrican.
The operational boundary of closed-loop sag compensation is defined by the response time of the hydraulic or servo accumulator and the frequency of measurable disturbances. If melt temperature varies rapidly because of an intermittent heater band fault, the sag controller may chase the disturbance shot by shot and introduce an oscillation in wall thickness. If the ambient airflow around the parison causes localised cooling on one side, the resulting asymmetric sag may not be detected by a single tip-position sensor. The drop test performance of a jerrican is sensitive to such asymmetries because the side-drop orientation may strike the cooled, thinned wall first. Instrumented production trials have shown that the spatial distribution of wall thickness is more important than the average wall thickness in predicting drop failures, but published data for the exact threshold under UN drop conditions is limited. For this reason, destructive testing of production samples is often performed every 2 h to 4 h on a designated drop test frame, using a filled jerrican and the appropriate UN height. The frequency is determined by the criticality of the product and the historical process capability of the line. A line producing packaging group I jerrycans with relative-density-adjusted heights above 2.0 m may require more frequent testing than a line producing packaging group III containers dropped from 0.8 m.
When a tooling change is introduced, such as a different blow pin diameter or a revised handle insert, the UN design type approval may require supplementary drop testing even if the base material and nominal wall thickness remain unchanged. The change alters the inflation ratio and the local draw that the sagged parison experiences, and it can shift the failure location from one orientation to another. Similarly, an increase in regrind ratio above the previously approved percentage may reduce melt strength and increase sag, even though the virgin resin lot is identical. The design type approval is tied to a specific combination of material, wall thickness, tooling geometry, and closure, and changes that affect any of these elements should be reviewed against the conditions recorded at the time of initial testing. Production batch release cannot rely solely on melt flow index and drop testing of occasional samples; it must include wall-thickness mapping, leakproofness testing, and control of parison length under actual moulding conditions. The leakproofness test specified in UN 6.1.5.2 and the hydraulic pressure test in UN 6.1.5.4 provide additional checks on sealing integrity, but they do not reproduce the impact-induced crack growth mode observed in the drop test. A jerrican that passes leakproofness may still fail the drop test if a thin sag-affected region reaches crack propagation under high strain rate. Conversely, a jerrican with a small leak path from a weak pinch-off weld can pass the drop test only if the weld remains intact under impact, which is precisely why the drop test remains the most severe discriminator of parison sag variation in UN-certified HDPE jerrycans.
