Is 0.5 kPa a figure from a standard?

Answer: No. Standards describe the method; the decay limit is agreed between maker and buyer and compares only within one cavity volume.

Should the case be empty or loaded during the test?

Answer: Empty. The test covers the shell and sealing pair; loading changes cavity volume and therefore the limit.

Why do readings rise in winter?

Answer: Temperature. Three degrees between shop and case is worth about 0.2 kPa; a two-hour soak or compensation corrects it.

Is a decay of zero the best result?

Answer: No. Zero can mean a blocked line, a dead sensor or a fixture squeezing the leak shut; prove the bench with a calibrated leak.

Air Leak Test by Pressure Drop: Reading the Result

Quick answer: The pressure decay method charges a case to 20 kPa, settles, then times 60 seconds of drop against a 0.5 kPa limit. A reading only counts when the case, the fixture and the sensor are all in order.

What an air leak test measures
Setting 20 kPa, 60 seconds and 0.5 kPa
How much temperature moves the number
Four places where leaks start
Passing the decay test is not IP67
FAQ

1. What an air leak test measures

An air leak test does not ask whether water gets in. It asks how much gas leaves a closed cavity per unit of time: leak rate equals pressure drop times cavity volume, divided by the measuring window. A 30 L case losing 0.5 kPa in 60 seconds passes roughly 250 Pa·L/s, enough to catch a missing gasket or a latch not fully closed.

The method spread through case plants for three reasons: no water contact, tens of seconds per piece, and full inspection instead of sampling.

The method has limits. An air leak test reports an equivalent leak, not a protection grade, and assumes the shell stays rigid. Below 2 mm wall thickness a lid sags under 20 kPa, cavity volume shifts, and the reading shifts with it.

2. Setting 20 kPa, 60 seconds and 0.5 kPa

Twenty kilopascals is not arbitrary. Lower and the signal drowns in noise; higher and a thin shell balloons, the lid sinks 1–2 mm, the gasket is pushed aside and a false leak appears. 20 kPa is about 2 m of water column, twice the head in IPX7.

Air leak test bench: tan rotomoulded case resting on a steel table, one draw latch at the left edge, moulded front handle, scuffed corners

The 60 seconds is best split in two. Gas heats as it is compressed, so an air leak test that times from the moment of charge reads a falsely high start pressure; settle 20 seconds, then count 60.

The 0.5 kPa line comes from volume. For one leak rate a larger cavity loses less pressure: 30 L dropping 0.5 kPa is about 250 Pa·L/s, while a 5 L case at the same figure holds one sixth of that leak, so small cases are often held to 0.1 kPa.

Sensor choice fails more often than the parameters do. A 0–35 kPa range at ±0.2% full scale with 1 Pa resolution is enough; below 1 Pa, a 0.5 kPa limit spans five counts and quantisation eats half the margin.

Parameter	Setting	Note
Charge pressure	20 kPa	about 2 m of water
Settle time	20 s	gas matches wall
Measuring window	60 s	counted after settle
Decay limit	0.5 kPa	30 L cavity
Sensor resolution	1 Pa	0–35 kPa range

3. How much temperature moves the number

Gas pressure follows temperature at roughly 0.34% per °C. At a 20 kPa charge one degree is 68 Pa, which sounds small until the budget is 500 Pa: a 3 °C difference takes 41% of it.

Difference	Drift at 20 kPa	Share of 0.5 kPa
1 °C	68 Pa	14%
3 °C	about 0.20 kPa	41%
5 °C	about 0.34 kPa	68%

That is why an air leak test calls for two hours of soak. A case straight off the moulding machine can run 8–10 °C above shop temperature, and on the fixture it reads a large decay and is condemned as leaking.

Three habits keep the error down: dry the air supply to within 2 °C of ambient, keep the bench off windows and outlets, and never hold the case against the body wall, since a 33 °C palm warms a patch fast.

Charging too quickly is the other trap in an air leak test. Filling 30 L to 20 kPa in 0.3 seconds can raise gas temperature past 5 °C, and a 20 second settle may not be enough.

4. Four places where leaks start

Eight out of ten failures in a lot come from four places, always in this order: relief valve, hinge bore, weld line, insert thread.

Air leak test aftermath: open tan case on a concrete floor, foam-lined lid, cut foam tray, four latches and grey shelving behind

The valve comes first: its sealing face is tiny. One grain of sand 0.5 mm across under the disc sends decay past 100 Pa while the case looks perfect.

The hinge bore comes second: it runs through the wall. Pin clearance is often 0.1–0.2 mm and a single O-ring does the sealing; pinched or missing at assembly, air goes straight through.

The weld line comes third, opposite the gate. Ten degrees low on melt or fifteen on mould temperature leaves a continuous micro-gap where two flow fronts meet, invisible by eye, obvious under pressure.

The insert thread comes fourth: blind-hole floor thickness. Tap 1 mm too deep, leave 0.8 mm of wall, and gas seeps along the thread.

Location	Cause	Fix
Relief valve	grit under the disc	blow out, retest
Hinge bore	O-ring pinched	new ring, add a stop
Weld line	low melt or mould temp	reset process
Insert thread	thin blind floor	deepen the floor

Validate the fixture first: clamp the halves, run 60 seconds dry, keep decay under 0.05 kPa, then feed a 0.5 mL/min calibrated leak and confirm the bench sees it. A spread above 0.1 kPa across three runs of an air leak test points at the bench, not the case.

5. Passing the decay test is not IP67

An air leak test uses gas; a protection grade uses liquid. Neither replaces the other. Surface tension lets a 0.05 mm hole hold back 1 m of water, while the same case shows the leak instantly under pressure.

That sensitivity cuts both ways: the decay method catches micro-leaks long before immersion, and also condemns sound parts, such as a gasket with release agent on its face, where gas travels the interface.

A case that passes IEC 60529 IPX7 at 1 m for 30 minutes usually decays under 0.3 kPa in 60 seconds; a case decaying 0.8 kPa shows droplets on the inner wall within 10 minutes.

The specification carries two lines: a decay limit with its parameters, and an immersion depth with its duration. One line merges two failure modes.

Keep the log wide enough too: date, serial, ambient temperature, set pressure, settle time, start pressure, end pressure, decay, verdict, bench number. Drop the ambient column and nobody can tell, months later, whether a lot drifted because it leaked or because it was cold.

Conclusion

A reading from an air leak test stands only when parameters, temperature, fixture and sensor all agree. Twenty kilopascals and 60 seconds set sensitivity, 0.5 kPa sets the verdict, and one degree of drift consumes a seventh of the budget. Report the decay figure beside an immersion result, and the question of whether a case is sealed gets an answer.

This article discusses only the container (case body) side of specifications and design intent, and does not cover any handling, loading or use of the contents.