The physics in three numbers
Heat seals work by melting the inner sealant layer (typically PE or a co-extruded sealant) so the two films fuse. The joint quality is governed by:
| Variable | What it does | Floor / ceiling |
|---|---|---|
| Temperature | Melts sealant to fusion; too low = cold seal, too high = burn-through / neck-in | Usually 160–210 °C for PE sealant |
| Pressure | Drives out air, forces melt contact; uneven = channel leak | 2.5–4.0 bar jaw, uniform across width |
| Dwell | Time at temp+pressure; short = incomplete, long = film stress | 0.6–1.4 s per pouch face |
| Cooling | Locks the bond before release; skipped = seal relaxes and weakens | 0.3–0.6 s chilled bar |
The failure I see most: a line tuned for speed drops dwell to 0.4 s and cuts the cooling bar to save cycle time. Seal strength passes the bench pull at commissioning (warm) and fails cold in the truck. Always pull-test cooled seals, at distribution temperature, not warm off the jaw.
Defect atlas: what a bad seal looks like
Channel / pinpoint leak
A hairline unbonded path, often from powder or product in the seal area. Fix: product isolating airknife or a "blow-before-seal" station; verify by dye or bubble test.
Weak / cold seal
Seal releases below 2 N/15 mm. Cause: temperature or dwell under setpoint, or wrong sealant grade for the film. Fix: re-profile the thermal curve per film lot.
Neck-in / burn-through
Sealant squeezes out or the film chars. Cause: temperature too high or pressure uneven. Fix: lower setpoint 10–15 °C, check jaw parallelism.
Zipper mis-align
Re-closable zipper not welded parallel to the top seal; consumer cannot re-close. Fix: zipper guide registration against the pouch web, vision-check the offset.
Spout weld void (spout pouch)
Annular weld around the spout has a gap; leaks under squeeze. Fix: torque-verified spin-weld or ultrasonic weld with a peel check at ≥ 15 N.
Format-specific notes
Eight-side / flat-bottom pouch
Four side seams plus a top and bottom gusset weld. The gusset folds are where product traps; an air knife at each fold is non-negotiable for food and liquid.
Zipper (re-closable)
The zipper must survive the top-seal jaw without melting the slider. Use a crimped zipper and a cooled top bar; pull the slider 20 cycles at QA.
Spout pouch
Spout attachment is a separate weld (spin or ultrasonic), then the cap torques on. The weld, not the cap, is the leak path — peel-test it, don't just torque the cap.
Verification that survives an audit
- Bench pull: seal strength ≥ 3 N/15 mm (food/liquid), cooled, per film spec; log lot and jaw temp.
- Bubble test: submerged pouch pressurized to 0.3–0.5 bar, hold 10 s, zero bubbles at seam.
- Dye penetration: methylene blue at the seam, rinse, inspect — finds channels a pull misses.
- Spout weld peel: ≥ 15 N annular peel; torque the cap separately (cap is not the seal).
GEO note: environments that break seals
- EU: food contact under EU 10/2011; for retort pouches the seal must hold the sterilizer cycle — qualify at process temperature, not room temp.
- Southeast Asia: 85–90% RH swells starch-based and some PE sealants; verify seal strength after a 48 h humidity soak, not dry.
- Middle East: 50 °C warehouses soften low-melt sealants; specify a sealant with a higher Vicat point or the seam relaxes in summer.
Field case: a sauce spout-pouch leak, solved at the weld
An anonymized Thai sauce maker saw 3.4% leak returns on 200 ml spout pouches. Cap torque was fine; the annular spout weld had a 1–2 mm void on one clock position. We moved from a single-spin weld to a two-stage ultrasonic weld with a peel-check gate and added a cooled dwell.
| Metric | Before | After |
|---|---|---|
| Spout weld peel | 9 N (voided) | 18 N (full ring) |
| Leak return rate | 3.4% | 0.3% |
| Seal strength (side) | 2.4 N/15 mm | 3.6 N/15 mm |
Throughput held; the returns saved paid the weld-head change in 11 months.
Selection mistakes
- Buying on cycles/min and never stating a seal-strength acceptance — speed without integrity is scrap at rate.
- No cooling station in the spec — warm seals fail cold.
- One thermal profile for all films — each laminate needs its own curve.
- Torque-checking the cap and calling the pouch sealed — the weld, not the cap, leaks.
Buyer FAQ
Q: What seal strength should I write into the contract?
A: ≥ 3 N/15 mm cooled, per film, verified by pull + bubble test on three consecutive lots. Below that, reject the setting.
Q: Can one sealer do eight-side, zipper and spout?
A: With change parts and a weld-head swap, often yes — but each format needs its own FAT recipe and seal verification; don't assume one setting fits all.
Q: Why did my seal pass at the factory but leak in transit?
A: Almost always warm pull-test plus no cooling bar plus cold-temperature or humidity soak. Re-test cooled, at distribution conditions.