Crazing: the defect you cannot see without a lamp

Polyethylene is tough, cheap and chemically resistant to a very wide range of things, which is exactly why it makes good containers. It has one weakness, and the weakness is specific: sustained mechanical stress plus a chemistry the polymer does not like.

Neither factor alone does much. A bottle under stress with water in it is fine for decades. A bottle full of an aggressive solvent with no stress on it degrades slowly. Combine them and the polymer chains slip apart along microscopic paths, and what you get is a craze: a network of hairline cracks, usually a few millimetres long, running in a pattern that follows the stress field.

Why you cannot see it

A craze is a void filled with drawn-out polymer fibrils. It has a refractive index close enough to the surrounding material that under diffuse light it disappears. Under a point source at a shallow angle it scatters light and lights up like frost on a window.

This is the whole reason it goes undetected. Buyers inspect containers outdoors in daylight, which is the worst possible lighting for the one defect that matters most.

How to check in fifteen seconds

1. 01Open the fill cap and put a torch inside, or shine one through the wall from outside on a natural-coloured bottle.
2. 02Move the beam so it strikes the wall at a shallow angle rather than straight on.
3. 03Look at the four vertical corners first. Stress concentrates there and crazing starts there.
4. 04Then look at the boss around the outlet, which is the second most common site.
5. 05Check the shoulders where the wall turns into the top, especially on containers that have been stacked.

The chemistries that cause it

Stress whitening is the sibling

Where crazing is chemical attack under stress, stress whitening is purely mechanical: the polymer has been strained past its elastic limit and the crystalline structure has been disrupted. It shows as pale, chalky patches, usually at corners or on the base where a container has been overloaded or dropped.

Both are terminal for a bottle in any application where you care about contents. Both are perfectly acceptable in a container that is going to hold rainwater at the back of a field. That distinction is why we grade rather than simply reject.

Why there is no repair

You cannot weld it, fill it or reinforce it. The cracks propagate under continued stress and the affected region is structurally compromised at a scale you cannot inspect. Anyone offering to repair a crazed bottle is offering to sell you a container that will fail later, somewhere less convenient.

What you can do is keep the rest. A crazed bottle in a sound cage is rebottling stock, and the frame underneath it is worth an order of magnitude more than the bottle ever was. That is not a consolation prize — it is the majority of the material by mass.

Preventing it in your own fleet

• Match the container to the chemistry. Aromatics and chlorinated solvents belong in stainless, not in polyethylene.
• Do not stack containers that are not rated for it, and never stack on a bowed cage.
• Avoid hot fills into polyethylene where you can. Heat plus stress accelerates everything.
• Store out of direct sunlight. UV embrittles the surface layer and lowers the threshold for everything else.
• Inspect with a lamp on receipt rather than on failure.

Why stress and chemistry have to occur together

Environmental stress cracking is a two-factor failure, and understanding that explains most of what looks arbitrary about it.

Polyethylene under tensile stress has polymer chains pulled taut in the amorphous regions between crystallites. A compatible chemistry migrates into those regions and reduces the energy needed for the chains to slide past one another. The result is that a load the material would carry indefinitely in air causes progressive failure in the presence of the fluid.

Remove either factor and the failure largely disappears. This is why an unstressed container of solvent degrades slowly and a stressed container of water does not degrade at all.

Where the stress comes from

• Hydrostatic load. A full container puts the lower wall under hoop stress, which is why crazing concentrates low rather than high.
• Geometry. Corners and the outlet boss are stress concentrators by shape, independent of load.
• Stacking. A container carrying another container has compressive load in the walls and tensile load at the corners.
• Moulding residual stress. Blow moulding leaves internal stress that never fully relaxes.
• Thermal cycling. Repeated expansion and contraction works the material.

Four of those five are unavoidable in normal service, which is why chemistry selection does nearly all the work in preventing crazing.

A practical inspection routine for a fleet

1. 01Inspect on receipt, not on failure. A torch and thirty seconds per container.
2. 02Re-inspect at every wash cycle, when the container is open and empty anyway.
3. 03Prioritise containers that have carried anything in the moderate-to-high risk categories.
4. 04Record findings against the serial so you can see whether a particular product is producing crazing across the fleet.
5. 05Retire on first confirmed crazing rather than monitoring it. It propagates and there is no repair.

What to do with a crazed container

Do not refill it, do not sell it for anything that will hold liquid under pressure, and do not throw the whole container away.

The cage underneath is almost certainly sound and is worth an order of magnitude more than the bottle. Sell it as rebottling stock or send it somewhere that will strip and reuse the frame. Roughly ninety-six percent of the steel in retired containers we handle is still perfectly good.

Questions about detection

Will a container with crazing leak immediately? Usually not, which is what makes it dangerous. Crazes propagate under continued stress and the failure typically arrives weeks or months later, often during handling rather than while standing.

Can you see it from outside? On a natural translucent bottle, sometimes, with a light behind it. On a black UV-stabilised bottle, no. Interior inspection through the fill opening is the only reliable method.

Does it affect a container used for water? Structurally yes, eventually. For a static rain tank on a stand that nobody moves, a lightly crazed bottle may serve for years. We still would not sell one for that, because the failure mode is a tank emptying itself without warning.

Is a hazy bottle crazed? Not necessarily. Uniform haze across the whole bottle is usually weathering or normal ageing of the polymer. Crazing is localised and follows the stress field — corners, the outlet boss, the shoulders.