Leak test vs leakage test terms
"Leak test", "leakage test" and "leakage testing" mean the same thing on a casting drawing: a check that no connected path crosses the wall under agreed conditions. What matters is not the word but the five values behind it: test medium, test pressure, method, dwell or sweep time, and the pass/fail limit. A leak test is not a porosity inspection. A casting with isolated internal pores can hold pressure, and a casting that looks dense can leak through a crack or cold shut that no X-ray view crossed.
We run two leakage tests in-house: water immersion (the bubble test) and helium leak testing with a sniffer. A casting is released as tested leak-free under the agreed pressure and method per drawing, never as "leak-proof" in the abstract.
Choose the method from the question
| Method | What it answers | Locates the leak | Limitation |
|---|---|---|---|
| Water immersion (bubble) | Is there a leak visible as a bubble stream at the test pressure? | Yes | Limited by what the operator can see; slow leaks can be missed |
| Pressure decay | Does the sealed volume lose more pressure than the limit in the test time? | No | Sensitive to temperature, volume and fixture leaks |
| Pressure (strength) test | Does the part withstand the proof pressure without failure? | Only gross leaks | A strength check, not a leak-rate measurement |
| Helium sniffer | Is there helium escaping at a point, below the visible-bubble threshold? | Yes | No total leak rate comparable to a vacuum-chamber test |
Pressure decay testing
A pressure decay test fills the sealed casting with air to the test pressure, isolates it, stabilises it and measures the pressure drop over a fixed time. The result is a number, which suits production control, but it says nothing about where the leak is. Part temperature, the internal volume and the fixture seals all move the reading, so the test pressure, fill and stabilisation time, measuring time and reject limit have to be stated together, and the rig proven with a known-good part.
When the drawing gives a leak rate in pressure-drop terms, pressure decay is the natural method; when the limit is tighter than decay can resolve, helium is the step up.
Pressure testing aluminum castings (air or water)
"Pressure testing" is often used for two different checks. A leak test at service pressure asks whether the medium can escape. A proof or strength test at a higher pressure asks whether the casting survives without deforming or cracking. Water is used for strength tests because it stores little energy; air is used for leak tests because it finds paths water cannot. State which one the drawing means, the medium, the pressure relative to service pressure, and the hold time.
The test pressure should normally be at or above the maximum the part sees in service. A test set below the service condition can pass a part that leaks at working pressure.
Helium leak testing
For tight leak-rate requirements we use helium sniffing: the casting is charged with helium and a probe connected to a mass spectrometer is swept over the surface to find the escape point. It detects leaks well below the visible-bubble size and locates them. Method selection, leak-rate units and report fields are covered on the dedicated page.
Helium leak testing of aluminum castings
Pre-test checklist, in order
Most false failures and false passes come from the rig, the seal-off or the record, not from the casting.
- Fix the criterion in writing. Medium, pressure, method, dwell or sweep time, pass/fail limit.
- Clean and dry the part. Oil, swarf and moisture can plug a fine leak path.
- Seal volumes not under test. Water must not enter dry cavities; helium must not escape through unrelated ports.
- Prove the fixture. Run a known-good part or blank first; a leaking rig is the most common false failure.
- Check calibration and background. Gauge accuracy at the set point; background helium before charging.
- Set the temperature. Part and water temperature affect readings and bubble visibility.
- Pressurise gradually and hold. A spike can open a path that does not exist in service.
- Observe the full dwell. Cutting observation short is a false pass.
- Record the location. "Bubble stream at the boss-to-wall junction" is a corrective-action input; "failed" is not.
- Dry and purge afterwards. Residual water corrodes; residual helium contaminates the next reading.
We design and build the leak-test fixture for each part in-house, so seal faces, clamping and seal-off are the same from the first part to the last.
What the report must state
There is no universal leak rate for every casting, and we do not invent one. Acceptance comes from the drawing or the agreed criteria.
- Criterion: medium, pressure, method, dwell or sweep time, limit
- Condition: part and medium temperature, sealing and masking
- Fixture used, so a repeat test is comparable
- Result with the leak point in part coordinates for any failure
- Traceability: part number, cavity or batch, date, operator
When the leak is not the casting
| Observed | Could be | How to tell |
|---|---|---|
| Leak at a drilled or tapped hole | Machining breakout, thread defect, or porosity the tool broke into | Inspect the hole wall at magnification: clean machined edges vs voids around the hole |
| Leak at a flange or mating face | Seal, face flatness, or casting behind the face | Replace the seal with a blank plate and re-test |
| Leak only above the test pressure | Path that opens at higher pressure | The criterion was below service pressure: review it, not the part |
| Leak in the rig before charging | Fixture, damaged seal, open port | A known-good part and a blank both fail: the rig |
| Helium signal, no bubble | Leak below bubble threshold, or background helium | Read background away from the part; persists with part vented = background |
| Bubbles that stop | Air trapped in a recess | Trapped air releases a few bubbles and stops; a leak continues through the dwell |
| Pass in test, leak in service | Crack opened by load or thermal cycling, or assembly damage | Outside a static test's scope; needs a separate verification plan |