The short answer
“No leakage” is not a test specification. Before tooling, define the assembled condition, test medium, pressure or vacuum, stabilization time, measurement interval, permissible leak rate or pressure change, fixture volume, sample frequency and record format. Use a locating method to find a path and a quantitative method to accept a part; do not treat the two results as interchangeable.
A pressure housing can pass a visual check and still fail after machining, plugging or gasket assembly. This guide shows how a buyer can turn an undefined leak concern into a repeatable RFQ and approval requirement for an aluminum die cast enclosure, valve body, motor housing or gearbox cover.
First draw the boundary that must remain sealed
Leak testing begins with a boundary, not with a machine setting. Mark every wall, machined face, threaded port, plug, insert, cover and gasket that separates the test volume from atmosphere or from a second circuit. A test performed on the raw casting answers a different question from a test performed after CNC machining and final assembly.
Record the exact part condition next to the requirement. Is the casting unmachined, machined, washed, coated, impregnated, fitted with plugs, or assembled with the production gasket and fastener torque? If that condition is left open, two factories can both report “pass” while testing different products.
Choose a method for the decision you need to make
A bubble test is useful for locating a visible leak path. Pressure-decay testing can support a measured acceptance decision when the fixture, volume, temperature and timing are controlled. Hydrostatic testing is relevant to components intended to retain liquids. Tracer-gas methods may be selected when a project requires greater sensitivity. None is automatically “best”; the method has to match the service risk and the specified limit.
| Method | Question it answers well | What must be controlled | Common misuse |
|---|---|---|---|
| Immersion or liquid-film bubbles | Where is a gross leak path? | Test pressure, dwell, wetting, orientation and observation | Reporting bubble size as an accurate total leak rate |
| Pressure decay | Does the closed volume remain within a defined loss limit? | Fixture volume, temperature, stabilization, sensor range and test time | Using the same time and limit for parts with different internal volumes |
| Hydrostatic | Does a liquid-retaining component show leakage under the stated condition? | Liquid, pressure, trapped air, dwell, drying and corrosion protection | Assuming it proves suitability for every gas-service requirement |
| Tracer gas | Can a smaller leak be detected or quantified at the required sensitivity? | Tracer concentration, calibration, background, fixture and units | Specifying helium without a justified limit or production plan |
ASTM E2930 describes pressure-decay practice and stresses that equipment, duration and technique must be chosen for the required sensitivity. ASTM E515 distinguishes bubble emission as a leak-locating, rather than total-leak-rate, technique. Those distinctions should appear in the buyer’s specification.
Eight fields that make a leak requirement quotable
The following fields belong in the RFQ even when the final values are still under engineering review. Writing “TBD by customer” is safer than allowing the supplier to invent a convenient default.
| Field | What to state | Why it changes the result or cost |
|---|---|---|
| Tested condition | Raw, machined, coated or fully assembled | Machining and assembly can open or close a leak path |
| Medium | Dry air, nitrogen, water, oil or specified tracer | Safety, sensitivity, cleanup and service relevance differ |
| Pressure boundary | Ports, plugs, chambers, gasket face and excluded volumes | The fixture must reproduce the intended circuit |
| Test level | Pressure, vacuum or differential, with units and tolerance | The same part can behave differently at another differential |
| Timing | Fill, stabilization, measurement and venting intervals | Temperature and material response can look like leakage |
| Acceptance | Maximum leak rate, pressure change or defined go/no-go observation | A result needs one unambiguous decision rule |
| Sampling | Tool trial, first article, per cavity, periodic or 100% | Fixture capacity and cycle time affect price and production flow |
| Evidence | Part revision, cavity, lot, equipment, calibration and result | Traceability makes failures diagnosable instead of anecdotal |
Separate real leakage from temperature and fixture effects
Pressure-decay results are sensitive to the complete test system. Compressed gas warms during filling and cools during stabilization. Flexible seals settle. A fixture or connector can leak even when the casting does not. Internal volume changes the relationship between a physical leak and the measured pressure change.
For sample approval, run a known-good master, a known-leak reference where the method requires it, and a fixture-only check. Lock the recipe to the part revision and identify the port-sealing method. If a production fixture is rebuilt or its volume changes, verify the measurement correlation before treating the new result as equivalent.
Connect a failed leak test back to casting and machining
A leak failure is not automatically a “porosity problem.” The route can pass through a shrinkage pocket, entrapped-gas porosity, a crack, an incomplete plug seal, a damaged gasket land, a thread root, a machining breakout or the test fixture itself. The fastest containment begins by preserving the failed part and locating the path before sectioning destroys the evidence.
- 01Repeat the test with the same recipe and verify the fixture
- 02Locate the path with an appropriate method
- 03Map the path to the cavity, casting feature and machining operation
- 04Review gates, overflows, vents, local mass and process records
- 05Check machining stock, tool path, plugs, seals and cleaning damage
- 06Confirm corrective action on new production under the original test condition
The companion guide on porosity control in aluminum die castings explains why critical zones should be marked before tooling. The CNC machining capability page shows how machined faces, bores and ports belong to the same approval route.
Two housings, two valid but different test plans
Electronics control enclosure
The principal boundary may be the cover joint and cable or connector ports. The engineering team must decide whether the requirement protects against process leakage under pressure, environmental ingress in service, or both. A factory pressure test does not by itself establish an ingress-protection rating; the assembly, seals and applicable product standard govern that claim.
Machined gearbox housing
The critical route may cross a machined seal face, bearing region, oil passage and installed plug. Testing before machining could miss a pore opened by stock removal. Testing before washing could allow chips to affect a plug or valve. The release plan may therefore require dimensional evidence plus a leak test in the finished condition.
Before approving the quotation, check these four interfaces
- Drawing to fixture: every tested port and excluded opening is identified on the current revision.
- Casting to machining: the porosity-sensitive zone, machining allowance and final sealing surface are reviewed together.
- Test to decision: units, recipe, acceptance limit, retest rule and handling of borderline results are written down.
- Result to lot: the report can be traced to part revision, cavity, production batch and any approved repair or impregnation route.
Use Huabo’s inspection overview to separate dimensional, material, appearance and functional evidence. For a comparable commercial scope, add the completed leak-test sheet to the die casting RFQ package.
Questions buyers ask about die cast housing leak tests
Should every aluminum die cast housing receive a 100% leak test?
Not automatically. Sampling follows the failure consequence, process capability, customer requirement and economics of the test. State the required frequency in the RFQ so fixture capacity and cycle time are quoted honestly.
Can CMM inspection prove that a housing will not leak?
No. CMM data can verify the geometry of faces, bores and hole patterns. Leakage can also involve material discontinuities, plugs, gaskets and assembly conditions, so a functional requirement needs an appropriate functional test.
Should vacuum impregnation be assumed as the normal repair?
No. Whether impregnation is permitted, prohibited or requires customer approval belongs in the controlled specification. It must not be used to hide an unstable casting process or an unapproved deviation.
Why can the same part pass one machine and fail another?
Different fixture volumes, seals, sensors, stabilization times, temperatures and calculation methods can produce non-equivalent results. Correlate equipment against the same masters and acceptance definition.
Technical references used for this guide
Use the current contract edition required by the customer. These sources explain method selection; they do not supply a project’s pressure, leak limit or sampling plan.
Technical review record
- Reviewer
- Huabo
- Reviewed
- 11 September 2026
- Evidence basis
- Huabo drawing-based casting and machining workflow; cited NADCA and ASTM method sources
- Scope boundary
- Customer drawings, specifications and written approvals remain authoritative
- Review trigger
- Recheck when the test method, product boundary, equipment, process route or cited standard changes
This page does not assign a universal pressure or permissible leak rate. Those values depend on the actual assembly, service risk and customer-controlled requirement.
Project-specific review
Send the housing condition—not only a “leak-free” note.
Provide the controlled drawing, ports, seal path, test medium, pressure or vacuum, timing, acceptance limit, annual demand and required record. Huabo will confirm the casting, machining and validation scope that can be quoted for the project.