What if all five steps find nothing?

Answer: Immerse the case 200 mm deep for 30 s and bubbles mark the leak. Blank the pressure valve first, or the valve mouth will bubble and mislead.

Answer: Yes. The line is 0.05 kPa and 0.06 sits above it; repeat three times, and if the median still exceeds the line the case fails.

A new gasket still fails. Where next?

Answer: Skip step one and go to latch preload and seam; a new strip compresses properly, so the leak is usually a latch that never loaded or a cold weld.

Why three readings instead of one?

Answer: A single decay carries temperature, gasket rebound and operator effects, with spreads up to 0.03 kPa. A median of three cuts misjudgement below 5%.

Repair or scrap after two attempts?

Answer: If more than two of gasket, weld and valve need work and labour passes 40 min, scrap it and keep latches, valve and handle as spares.

Case Seal Failure Diagnosis: Five-Step Airtight Diagnosis

Quick answer: A case seal failure diagnosis starts at one number: hold 0.3 kPa for 60 s and a decay above 0.05 kPa fails the case, so work gasket, valve, latch, seam and ports in that order.

Reading the decay before opening anything
Step one: appearance and gasket compression
Step two: valve seat return
Step three: latch preload
Step four: shell and seam
Step five: ports and plugs
FAQ

1. Reading the decay before opening anything

A case that passed at the end of the line returns with a 0.12 kPa decay, more than twice the limit. Do not open it: re-pressurise to 0.3 kPa, hold 60 s, read again. A repeatable figure means a stable leak; a figure that changes every run means the gasket was pinched at closing.

Only the difference over the 60 s counts: starting at 0.32 or 0.28 kPa changes nothing, and a 0.4 class gauge on a 2 kPa span reads 0.05 kPa at a glance.

Temperature writes false readings: a case moved from a 5 ℃ store into a 25 ℃ bay needs 30 min to equalise, and the gauge drifts 0.02 kPa meanwhile.

Repeat three times, venting to zero between runs with 2 min gaps; a spread above 0.03 kPa points at the fixture rather than the case.

Decay in 60 s (kPa)	Verdict	Next move
0.02 or less	Pass	Record and release
0.02–0.05	Borderline pass	Repeat three times, take the median
0.05–0.15	Fail	Enter the five steps
Above 0.15	Gross fail	Check whether the gasket left its groove

The fixture leaks too: aged O-rings on couplers decay on every case, so blank the line with a solid plate; under 0.01 kPa means the reading belongs to the case.

2. Step one: appearance and gasket compression

Four minutes, two questions: is the gasket in its groove, and how far is it compressed. A closed lid should leave the strip 3.15 to 3.40 mm high, 25% to 30% compression and the first guarantee of tightness.

Case seal failure diagnosis samples: two black shells open on a pale shelf, orange and black cut foam inside, a black case below

The arithmetic is short: a 3.2 mm groove with a 4.5 mm strip must lose 1.1 to 1.35 mm at the joint. Check eight points with calipers; any point above 3.40 mm never compressed, and a failed diagnosis lands there.

A strip bulging out of its groove was never seated at assembly, and a flat spot means it was trapped the last time the lid shut; a gloss line at full height means the leak is elsewhere.

Joint evenness says more than any single reading: a gap varying over 0.5 mm means a skewed lid, and a strap offcut holding the strip 0.3 mm off the bottom is cleared by compressed air, not a new gasket.

3. Step two: valve seat return

Three minutes on the pressure-equalisation valve: the disc is 14 mm across, opens near 0.3 kPa, and must sit back on its seat, or the case carries a 14 mm hole.

Press the disc for 2 s and release: it should seat within 1 s with a light click. A slow return, a visible gap, or a 0.05 mm feeler under the edge all fail, and a spring that has lost 1 mm of free length is replaced rather than adjusted.

Soap solution is the judge: coat the mouth, pressurise to 0.3 kPa and watch 30 s; clustered bubbles mean grit between disc and seat, one large bubble a distorted disc.

The threaded joint between body and shell stays invisible until disturbed: grip the body and try to turn it, and movement means it was never torqued, so re-tighten and re-soap.

4. Step three: latch preload

Five minutes across four latches, each travelling 6 mm and, fully closed, loading the lid by 1.0 to 1.5 mm of interference; less means a lid shut but not closed.

Case seal failure diagnosis shelf holding four open cases, blue and red lids raised, metal fittings in red foam, grey tray in the blue case

Test with the pull-gap method: with every latch down, pull the lid joint outward. A gap of 0.3 mm or less passes; at 0.5 mm the latch rests on the keeper, the strip never reaches 25% compression and the check fails.

Three sources explain a latch that will not load: a loose base, a worn hook, or a lid deformed in transit, so re-torque the base, inspect the hook and check the lid against a surface plate.

Close the long-side latches first and the end latches last, so displaced air has a path out. A steady 0.03 kPa decay after every latch is down usually means a gap under the base, which a 0.5 mm shim normally closes.

5. Step four: shell and seam

Five minutes on shell and seam, the leak most often misread as a gasket problem. Most cases are two mouldings joined on a 3 mm seam sealed by ultrasonic welding or adhesive, and any unwelded stretch leads straight to a failed airtight check.

Work in segments: tape off a stretch, pressurise to 0.3 kPa, soap for 30 s. Segments of 200 mm mean eight stops around a 1.6 m perimeter.

Bubble shape separates the two causes: a cold weld gives a fixed line of clustered bubbles, a shell pore a single bubble in a random spot. Cavities between liner and shell carry air sideways, so a liner-edge leak surfaces centimetres away and the edge needs soap as well.

Start at the hinge side, which takes the highest load every time the lid closes and carries the highest share of cold welds.

6. Step five: ports and plugs

Three minutes on every point that penetrates the shell, the smallest part of a case seal failure diagnosis.

All seal on an O-ring or grommet lip 1 to 2 mm wide, so slight eccentricity leaks; soap each at 0.3 kPa and watch 30 s, and the bubble marks the leak without a strip-down.

Port	Check	Pass criterion	Time (min)
Pressure valve	Press disc, release, soap	Seats within 1 s, no bubble in 30 s	1
Threaded plug	Re-torque to 1.5 N·m, soap	No bubble in 30 s	1
Cable grommet	Lift lip and inspect	No rolled lip, compression 25% or more	0.5
Relief port	Press three times, soap	Normal rebound, no bubble	0.5

Plugs fail on torque: 1.5 N·m suits an M12 plug on a 2.5 mm O-ring, and either too little or too much ends in a failed check. Walk the ports from the drawing, or one missed plug turns a 20 min location into 40 min of rework.

Conclusion

The five steps matter less as a list than as five judgeable actions inside 20 min, each ending in a number or a feel, ordered by leak share: gasket, valve and latch carry about 70%. Once all five are done and the case still fails, three outcomes remain: repair it when strip and latches measure in tolerance and the leak sits in a seam or port that opens; downgrade it when leaks scatter and the structure no longer suits its rating; scrap it when the shell shows a through-crack or several cold welds and labour passes 40 min.

Step	Criterion	Time (min)
1. Appearance and gasket	Strip 3.15–3.40 mm at the joint, 25–30% compression	4
2. Valve return	Seats within 1 s, no bubble at 0.3 kPa in 30 s	3
3. Latch preload	Pull gap 0.3 mm or less	5
4. Shell and seam	Segmented taped test, no continuous bubbles in 30 s	5
5. Ports and plugs	Soap each port, no bubble in 30 s	3
Total	Leak located to a single point	20

This article discusses container-side specifications and design implications only, and does not cover any handling, loading or use of the contents.