Rotational moulding of linear low-density polyethylene (LLDPE) intermediate bulk containers (IBCs) for dangerous goods transport typically produces a one-piece seamless shell classified under UN 31H1 with nominal wall thickness between 5 mm and 7 mm. Production-scale machines consist of three- or four-arm carousels with independent biaxial rotation, mould cavity surface temperatures between 260°C and 300°C, and demoulding when the part sidewall reaches 70°C to 90°C. Critical failure zones under the UN drop test are the top filling neck area and the lower valve boss, where local wall thickness may fall to 40% to 60% of the nominal design value because of powder bridging, corner radius limitations, or venting constraints. These zones also exhibit high residual hoop stresses because the outer wall freezes first while the inner wall remains molten, generating a through-thickness density gradient, crystallinity gradient, and differential shrinkage. The laboratory evaluation of impact strength must therefore separate material ductility from geometric stress concentration; methods include instrumented puncture per ISO 6603-2, falling-dart impact per ASTM D5628, tensile impact per ISO 8256, and full-scale UN drop tests on preconditioned IBCs at -18°C. Environmental stress crack resistance is normally evaluated by ASTM D1693 bent-strip coupons immersed in 10% Igepal CO-630 solution at 50°C or 100°C. A significant limitation is that the bent-strip coupon imposes a single surface strain state that does not reproduce the biaxial residual strain field at an IBC corner transition. Consequently, rotomoulders maintain dual datasets: small-coupon ESCR for lot-to-lot material acceptance and full-scale container drop tests for design-type qualification and periodic verification.

What Is the Governing Influence of Tie-Chain Density on Low-Temperature Dart Impact in Rotomoulded LLDPE IBCs?

The low-temperature impact response of a rotationally moulded LLDPE shell is not a single material property; it is a convolution of copolymer architecture, crystallisation history, and the local multiaxial strain rate applied during a drop event. In an octene-based LLDPE, the hexyl branches depress the average lamellar thickness and increase the probability that a high-molecular-weight chain will traverse more than one lamella, thereby forming a load-transferring tie chain. The tie-chain density rises as the weight-average molecular weight increases and as the short-chain branching distribution narrows. Dart impact values at -40°C may range from 120 J to 220 J for commercial octene LLDPE rotomoulding grades with density 0.935 g/cm³ to 0.945 g/cm³ and melt flow index 3.0 g/10 min to 6.0 g/10 min at 190°C/2.16 kg, while butene-based equivalents frequently fall below 80 J at the same temperature because the shorter butyl branch allows thicker lamellae and fewer tie molecules. The ductile-brittle transition temperature measured by ASTM D5628 on rotomoulded plaques is typically 15°C to 25°C lower for octene grades than for butene grades of equivalent density and melt index. However, those plaque values are not directly transferable to a 1,000 L IBC because the drop-test strain rate at a sharp fillet can exceed the rate at which lamellar slip can dissipate energy; the local transition therefore shifts upward by as much as 10°C to 20°C. For this reason, design qualification at -18°C requires not only high tie-chain density but also a minimum practical corner radius of 12 mm to 20 mm and controlled powder size distribution below 500 µm. Rheological data show that low-shear viscosity and zero-shear viscosity increase with molecular weight while the onset of shear thinning shifts to lower angular frequency; a rotomoulding grade with a zero-shear viscosity above 2,500 Pa·s at 190°C may exhibit insufficient bubble release and pinhole formation, even though its ESCR and impact resistance are superior. Thus the selection of a high-performance octene LLDPE is constrained at the upper end by powder consolidation and surface aesthetic requirements.

The tabulated ranges are representative of non-cross-linked LLDPE rotomoulding grades from polyolefin producer technical bulletins. Published data for specific IBC geometry is limited, and full-scale verification remains necessary because plaque-level values do not capture the multiaxial stress state at container transitions.

A recurrent quality-control discrepancy in 31H1 IBC production arises when a laboratory ASTM D1693 lot demonstrates F50 greater than 1,000 h in 10% Igepal CO-630 at 50°C, while field containers show branched surface cracks after 24 months of cyclic hydraulic loading. The discrepancy is explained by three factors. First, the bent-strip specimen is machined from a compression-moulded or rotomoulded flat plaque where cooling is relatively uniform; the IBC corner zone has a crystallinity gradient from the outer to inner surface of up to 12% as measured by differential scanning calorimetry. Second, the Igepal test fluid does not match the swelling behaviour of certain aliphatic esters, quaternary ammonium disinfectants, or oxidizing hypochlorite solutions that plasticise the amorphous tie-chain regions and lower the stress-crack activation energy. Third, the residual hoop stress in a demoulded IBC can reach 2 MPa to 4 MPa at transition fillets, whereas the bent-strip fixture imposes a controlled outer-fibre strain of approximately 1.5% to 2.0%. For compatibility screening, producers therefore run supplementary immersion tests per ASTM D543 on loaded flexural specimens or whole-container patch coupons in the intended chemical medium at 40°C to 60°C for 500 h to 1,000 h, then retest residual tensile elongation and instrumented puncture energy. The use of octene-based LLDPE with a density below 0.940 g/cm³ generally improves slow crack growth resistance because the greater amorphous fraction permits more viscous energy dissipation, but it also reduces the top-load buckling resistance and the bursting pressure margin under hydraulic test. These competing requirements force a density window that is normally 0.936 g/cm³ to 0.942 g/cm³ for non-cross-linked single-wall IBCs.

Oxidative Degradation During Oven Dwell Produces a Thickness-Dependent Property Cliff

In a rotational moulding cycle, the inner surface of an LLDPE IBC remains at near-peak temperature for substantially longer than the outer surface because the outer layer first contacts the heated mould wall and also cools first when the mould enters the forced-air cooling station. On a four-arm carousel with an oven dwell of 30 min to 45 min, the inner wall may remain above 190°C for 15 min to 20 min while exposed to residual oxygen in the mould cavity. This exposure consumes hindered phenolic antioxidants through hydrogen donation, and once the local antioxidant concentration falls below a critical level, the rate of carbonyl formation accelerates. Oxidation induction time measured by ISO 11357-6 or ASTM D3895 at 200°C may decline from an initial 25 min to 40 min on virgin powder to less than 5 min on the inner surface after a single extended cycle. The property cliff is thickness-dependent because the degraded inner skin may represent 10% to 20% of the wall. Instrumented puncture at -18°C on plaques with a degraded inner skin often shows a 30% to 50% reduction in total energy absorption compared with plaques stabilised to retain OIT above 20 min. Fourier-transform infrared spectroscopy on microtomed sections reveals a carbonyl index increase from below 0.05 to above 0.15 in the affected layer, while the colour shifts to a yellowness index above 2 per ASTM E313. The critical processing control is not oven set-point alone but peak internal air temperature, which should be kept below 205°C and ideally between 190°C and 200°C for a non-cross-linked LLDPE. When PIAT exceeds 210°C, gas bubble coalescence at the inner surface creates pinholes that act as crack-initiation sites during hydraulic pressure cycling. A production work instruction for 31H1 containers should therefore specify a PIAT tolerance of ±5°C, not the ±15°C often tolerated in less safety-critical rotomoulded parts. The stabiliser package must also be selected for absence of amine-based antistatic additives when the IBC is intended for oxidizer service, because amine additives can accelerate oxidative degradation and are incompatible with hypochlorite re-use cycles.

Environmental Stress Crack Resistance, Copolymer Architecture, and 31H1 Design-Type Qualification

The 31H1 design-type approval for rigid plastics IBCs requires a sequence of performance tests on representative containers, including drop testing at -18°C after conditioning with 100% filling material, hydraulic pressure testing to the relevant gauge pressure, leakproofness testing, and stacking for 28 days at 40°C. These tests are specified in the UN Manual of Tests and Criteria and adopted in ADR/RID and 49 CFR. Yet none of these design-type tests directly measures ESCR; they capture only the initial structural integrity of the moulded article. Consequently, for service fluids that are known secondary stress-cracking agents, the shipper must conduct compatibility testing under ADR 4.1.1.5 or equivalent national clauses. The ESCR ranking of an LLDPE IBC grade is governed by the type, content, and distribution of the butene, hexene, or octene comonomer. Octene-based LLDPE has the longest sequence of methylene units between short-chain branches, which permits a higher frequency of stress-transmitting tie chains and resists lamellar disentanglement. In bent-strip testing per ASTM D1693, condition B at 50°C, a medium-density rotomoulding grade with density 0.938 g/cm³ and MI 4.0 g/10 min may exhibit F50 values of 150 h, 400 h, and >1,000 h for butene, hexene, and octene copolymers respectively. The differences become larger at 100°C because the test temperature approaches the alpha-relaxation region where long-chain branching and molecular weight distribution dominate the disentanglement kinetics. The compatibility matrix should also include the user’s duty to avoid storing hydrofluoric acid, strong oxidizing acids above specified concentrations, and certain chlorinated solvents unless specific published compatibility data exist.

The hydraulic pressure test is not a material property test but a system verification of weld-free moulding integrity. For rotomoulded IBC shells, the common leak path is not through the wall but along the valve flange or fill cap closure. However, a stress crack can initiate at the internal surface of the bottom sump if the part is demoulded with a high frozen-in strain and exposed to a wetting agent. This is why full-container ESCR screening is sometimes conducted by pressurising containers at 20 kPa to 30 kPa and maintaining them at 40°C for 14 days in contact with the intended chemical. Published data for whole-container ESCR compatibility with all possible chemical mixtures is limited; this matrix does not substitute for the shipper’s compatibility demonstration.

The effect of post-industrial regrind incorporation on IBC impact and ESCR is not linear and cannot be approximated by a simple rule of mixtures. Reprocessing a rotomoulding-grade LLDPE through a single-screw compounder with a 30:1 L/D ratio and pelletising at 190°C to 220°C consumes additional antioxidant, shortens the molecular weight distribution, and introduces low levels of oxygenated species that accelerate slow crack growth. When 15% regrind is re-extruded into powder and blended with virgin material, the loss of ESCR may be less than 10% if the original powder had a high OIT reserve and the grinding step is cooled. At 50% regrind, however, F50 values can decline by 40% to 60% and the low-temperature dart impact may drop below 100 J at -40°C. The failure mode is often not visual contamination but gel formation from crosslinked or degraded skins near the inner surface. A conservative specification for 31H1 production limits regrind to 20% of the total weight, excludes regrind from any part of the container below the liquid level, and requires melt-flow ratio monitoring per ASTM D1238 using the 21.6 kg load to detect incipient long-chain branching. The ESCR of a regrind-containing wall also depends on the particle size distribution after grinding; particles coarser than 425 µm may not consolidate at the inner surface and leave microvoids that act as stress-raisers during hydraulic cycling. In production, powder with a median particle size of 250 µm to 350 µm and a maximum of 500 µm provides sufficient melt densification for a 6 mm nominal wall without excessive oven time. Powder should be tumbled with dry pigment before loading, and moisture must be kept below 0.05% to avoid steam pinholes.

When Tetrahydrofuran-Based Cleaning Fluids Replace Alkaline Hypochlorite Solutions in Reused IBCs

Reconditioned IBCs are frequently washed with alkaline hypochlorite solutions at 60°C to 80°C for tank sanitizing, but some closed-loop chemical users replace the hypochlorite wash with tetrahydrofuran-based or glycol ether-containing cleaning fluids to remove polymerised residues. This substitution creates a severe environmental stress cracking environment because the solvent plasticises the amorphous phase and lowers the craze yield stress below the residual hoop stress in the wall. In ASTM D1693 screening, a replacement fluid may show no failure for 1,000 h at 50°C, yet a full IBC under 20 kPa internal pressure can show branched cracks within 200 h when the cleaning agent is left in contact with the floor sump. The difference arises from the notch geometry: the bent strip has a defined razor notch, whereas a rotomoulded inner surface contains microscopic pinholes, powder grain boundaries, and stress whitening zones that behave as pre-existing flaws. Resistance in solvent-containing service is controlled by the Hansen solubility parameter distance between the solvent and the polyethylene, but more importantly by the ability of the solvent to diffuse into the tie-chain interphase. Small polar molecules such as tetrahydrofuran have a high diffusion coefficient in the amorphous layer because their molar volume is small, while their polar component is sufficient to disrupt weak intermolecular contacts. The result is a reduction in failure energy measured by instrumented puncture after 72 h immersion of 10% to 25% in a 6 mm plaque. A safe practice for reconditioned IBCs is to restrict solvent-based cleaning to the outside surface, to rinse with demineralised water at 60°C, and to dry the interior with filtered air at 40°C for at least 8 h before returning the container to service. Any use of tetrahydrofuran at concentrations above 1% in the contained liquid should be considered outside the qualified compatibility envelope unless the specific grade has published failure-time data from a whole-container test.

The cooling step after oven dwell is as decisive as the heating step. When a rotomoulded IBC is removed from the oven and subjected to forced ambient air, the outer surface cools at rates between 15°C/min and 40°C/min, whereas the inner surface cools at 5°C/min to 15°C/min. This differential cooling produces a spherulite size distribution from 2 µm to 8 µm near the outer wall to 8 µm to 20 µm near the inner wall, with larger spherulites having weaker boundaries and higher susceptibility to slow crack growth. Quenching the entire mould in water can reduce spherulite size and improve impact strength but increases residual stress and warpage, especially around the top cap and fork-entry pockets. The optimum cooling protocol for LLDPE IBC shells is a two-stage forced-air profile: 8 min at 30°C with high airflow, followed by 12 min at ambient still air, before demoulding at 70°C to 85°C. If the mould is opened too early, the part shrinks unevenly around the core pins and can develop microcracks at the valve boss. If it is left too long, cycle time increases without property benefit.

Full-scale IBC drop testing at -18°C remains the only reliable predictor of impact resistance in a rotomoulded geometry, but it is expensive and statistically noisy. A 1,000 L IBC filled with water/glycol and dropped from 1.2 m onto a rigid steel plate according to UN design-type testing shows local failure when the strain energy at the impact corner exceeds the thickness-modified energy absorption of the polymer. Instrumented container drops with triaxial accelerometers and high-speed video indicate that the first failure event often occurs within 5 ms of contact, with peak strain rates at transition fillets exceeding 10 s⁻¹. Under those rates, the LLDPE still deforms by shear yielding if the temperature is above the ductile-brittle transition, but the local adiabatic heating can be insufficient to prevent crack propagation when the wall thickness is below 4 mm. This is why a minimum measured wall thickness of 4.5 mm at all points below the liquid level is specified by many 31H1 approvals, even when the nominal design thickness is 6 mm. The correlation between laboratory ESCR and full-scale drop impact is weak in the short term; ESCR is a slow-crack parameter, whereas drop impact is a high-speed failure parameter. They are unified only by the need for sufficient tie-chain density and by the avoidance of degraded inner surfaces. A container that displays stress whitening after a drop test but no visible crack may pass the UN test and yet fail later in service because the whitening zone is a precursor to environmental stress cracking if the exterior is exposed to a liquid that wets the scratched surface. This operational boundary is particularly important for IBCs used in dairy sanitizers, agricultural chemicals, and metalworking fluids.