Engineering5 min read

A closure rarely fails on its own. It fails in combination — with a neck finish that drifts, a capping head that applies more torque than the design assumed, or a product that behaves differently at forty degrees than it did in the laboratory.

The neck finish is the usual suspect

Nominal sizes hide real variation. Two 28 mm finishes from different suppliers can differ in thread height, root diameter and the flatness of the sealing land. A closure designed against a drawing rather than a measured sample inherits that variation as a leak path.

Torque is a window, not a number

Every closure has a torque range where it seals and still opens by hand. Set the capper below it and the pack leaks; above it and the liner distorts or the band shears. That window narrows as the closure ages, so it should be measured after conditioning, not only on fresh parts.

The product is part of the seal

Surfactants creep along surfaces. Solvents swell polymers. Hot filling leaves a vacuum that pulls the panel inward as the contents cool. None of this shows up in a room-temperature water test, which is why compatibility trials have to run with the real formulation.

What to specify instead

Give your supplier the measured neck finish, the capper settings, the fill temperature and the actual product. A closure specified against those four will survive the line. One specified against a catalogue number is a gamble that usually gets settled during a production run.