You set application torque at the capper, audit removal torque on 30 bottles, everything reads in band — then a downstream sample or a customer return comes back with a loose cap or lost fizz. Before you suspect the capper, look at what the bottle passed through in between. If your line runs a warmer after filling (common on carbonated and cold-fill lines, to lift the bottle above dew point so it does not sweat before labelling), the cap you tightened cold is not the cap the consumer opens. Retained torque falls across the warmer — and that drop is physics, not a fault.
Why the torque falls
A screw cap seals because its thread pulls down against the neck thread and squeezes the sealing element. That clamp load depends on the interference between two plastic parts, and both change in the warmer:
- Stress relaxation — the main driver. Plastics under load shed stress over time, and far faster when warm. Much of the hoop and axial stress locked into the cap and neck at capping relaxes during the warmer’s dwell, and does not fully recover on cooling. The pack leaves the warmer with permanently less retained torque than it had at the capper.
- Differential thermal expansion. The polyolefin cap (PP or HDPE) and the PET neck expand at different rates as the pack heats, shifting the thread and sealing-surface engagement. The net direction depends on the seal type, but for most top-seal and thread-friction designs it slackens the interference that generated your torque.
So removal torque measured after the warmer is typically lower than the cold reading right after capping — which is already below the application torque you dialled in. Nothing is broken; the pack relaxed. The real question is whether it relaxed past the point where the seal still holds pressure and the cap still opens cleanly.
What it means for setting torque
The classic mistake is to tune application torque so the cold, at-capper removal reading sits mid-band for openability. If the line then runs a warmer, that number drifts down and can cross below the seal-holding threshold — and on a carbonated pack the first symptom is slow CO₂ loss and a flat product weeks later, not an obvious leak on the line. Validate against the after-warmer reading, not the capper reading. Delta publishes an application-torque window per closure family on the caps & closures page; in a warm plant, target the middle of that window rather than the upper bound, leaving room for relaxation without over-tightening.
What to monitor
- Removal torque before and after the warmer — run the same audit at both points to see the real drop on your line (the 30-bottle method).
- Warmer outlet temperature and dwell — hotter and longer means more relaxation.
- Cap and neck-finish dimensions — sealing surface, thread and tamper-evidence band must be in tolerance before the warmer effect even matters.
- Leak / seal test — the outcome that counts; pair the torque audit with a pressure or vacuum check (closure leakage troubleshooting).
Where the closure comes in
Relaxation is a property of the loaded parts, so the closure design sets your margin: material (closure material guide), liner or sealing-fin geometry, and the fit to the neck finish all decide how much retained torque survives the warmer. If your line runs a warmer, say so at RFQ — Delta El Nile for Industry specs the closure and confirms the neck-finish fit so the seal holds after it cools, not just at the capper. Find your closure or build an RFQ.