Case Study - Polyethylene Container Failure

Smithers Rapra were requested to provide a report on a failed HDPE blow moulded container that was the cause of dispute between the blow molder & the wholesaler.

History

Five litre HDPE blow moulded containers had a long & successful history as a means of short-term storage of a particular agrochemical. Typically the containers would be pallet stacked in a warehouse environment for several months, pending distribution to retail outlets.

One consignment of a thousand containers started to leak within two weeks of storage. Leaking containers were buckled as shown in Figure 1. The agrochemical in concentrated form was known to be a health hazard, so specialists were deemed necessary to decontaminate the warehouse.

The wholesaler sought recompense from the container supplier. The supplier responded by claiming that the recommended 'top load' on the containers had been exceeded due to overstacking. The buckling of the containers indicated that this was the case. The wholesaler knew that the top load limits had not been exceeded but could not prove this to the satisfaction of the supplier. Feeling confident that they would be vindicated they agreed to accept the findings of an independent expert specialising in stress analysis.

Whilst awaiting the expert's report, consignments failed in a similar manner in 5 warehouses in other parts of the country. Poor storage practice was relegated as a candidate cause.

Inspection & Analysis

From a plastics technologist's perspective, the most significant feature of the failures was the presence of microcracking in all containers that had buckled & leaked. Figure 2 clearly reveals such cracking on the tensile surface of a buckled fold. The 'linear & stepped' appearance of the microcracks is strongly indicative of environmental stress cracking.
However, from the stress analyst's perspective the cracking was of only secondary interest. Common sense & 'intuition' would suggest that the cracking was caused by the buckling.
High powered microscopy revealed the presence of extensive microcracking on the inside surface of all containers including those that had not buckled.

Failure Diagnosis

The ESC resistances of material batches were compared using a simple bent strip test. Material from the latest & therefore suspect batches were found to be significantly inferior to material from old batches. Under equal strain, time to crack in the agrochemical was reduced by more than a factor of 10.
On being presented with this evidence, the blow moulder (ISO 9000 accredited) admitted a change in material supply. He had been offered the 'same grade' in terms of density & melt flow index, at a cost saving of 10%. He was unaware that the premium grade had originally been chosen & formulated for its excellent stress crack resistance (e.g. lightly copolymerised with an enhanced high molecular weight fraction).

The causative connection between ESC & buckling is apparent from Figure 3. This compares the creep behaviour of a plastics material in air & in a stress cracking environment. Prior to craze or crack initiation the time dependent stiffness of the material is unaffected by the environment. After initiation the creep rate accelerates & the time to buckle under constant top load decreases from T (in a passive environment) to T* (in an aggressive environment).

Lessons & Consequences

1. ISO 9000 & equivalent systems promote but do not ensure the continuity of quality.
2. Apparently small differences in material grade can significantly affect durability.
3. Misdiagnosis is common. Had there been no further consignment failures an incorrect diagnosis would have been made & the wrong party blamed.
4. The blow moulder was insured against consequential losses up to £250,000. This limit was fully exploited in the subsequent settlement.