Leak Testing for Critical Components

Manifolds, housings, and ducts are deceptively hard to qualify. They often combine thin walls, complex internal passages, multiple ports, and welded/brazed features—exactly the geometry that can hide porosity, microcracks, or assembly leaks until late in the program. In defense and aerospace, a “small” leak can drive mission failure, contamination, safety risk, or expensive rework after downstream integration.

For buyers, the challenge is that “leak test” can mean anything from a quick bubble check to a calibrated helium mass spectrometer test with quantified sensitivity. If you are sourcing leak testing services for critical hardware, the best outcomes come from specifying the method, acceptance criteria, part condition, and documentation package up front—especially when parts are produced via additive manufacturing (AM) like powder bed fusion (PBF) DMLS/SLM, or densified via HIP / PM-HIP and finished with CNC machining.

This article breaks down what to ask for so engineering, procurement, and program teams can align on a test that is technically valid, repeatable, and audit-ready under regulated workflows (AS9100, ITAR, DFARS, NADCAP where applicable).

Common leak test methods

The “right” leak test depends on the fluid system, allowable leak rate, part geometry, and whether you need a quantitative result (measured leak rate) or a qualitative screen (leak/no leak). For critical manifolds and housings, prioritize methods that produce a numeric leak rate with known sensitivity and calibrated equipment.

1) Helium mass spectrometer leak testing (MSLT)

Helium testing is widely used for aerospace hardware because it is sensitive and can be configured as either vacuum or pressure (“sniffer”) testing.

How it works (typical workflows):
Vacuum method: The part is placed in a vacuum chamber (or connected to a vacuum fixture). Helium is applied to the outside (spray) or to one side of the part, and the mass spectrometer measures helium that passes through leak paths into the evacuated volume. This is generally the most sensitive configuration for small leak rates.
Pressure method (sniffer): The part is pressurized with helium (or a helium mix), and a probe “sniffs” around joints, seals, welds, and suspect features to localize the leak. Sensitivity is lower than vacuum mode but is useful for troubleshooting and rework validation.

Where it fits: High-value components, safety-critical housings, flight hardware, and applications requiring very low leakage (e.g., pneumatic systems, propellant-related hardware, inert gas systems, or contamination-sensitive electronics enclosures).

2) Pressure decay / pressure drop testing

Pressure decay uses a known test volume pressurized with air, nitrogen, or another gas, then measures pressure change over time. It is common in production because it is fast and automatable, but it can be less sensitive than helium mass spectrometry and is highly dependent on temperature stability and volume estimation.

Key practical point: A pressure decay result is not “just pressure over time.” The supplier must control and report stabilization time, test duration, temperature compensation (or controlled environment), and total internal volume to convert decay to an equivalent leak rate.

Where it fits: Medium-to-high volume programs, robust leak limits, and components with larger acceptable leakage. It is also effective as an in-process screen before final assembly or coating.

3) Bubble / immersion testing

Bubble testing pressurizes the part (often with air) and submerges it in liquid to observe bubbles. It is simple and can find gross leaks, but it is subjective, strongly dependent on operator technique, and poor at quantifying low leak rates.

Where it fits: Noncritical hardware, early process development, weld/braze screening, and troubleshooting when the acceptance criterion is “no visible bubbles” (which should be treated as a qualitative screen, not a quantified leak rate).

4) Flow testing

Flow testing measures flow rate through the part at a specified differential pressure. For manifolds and ducts, this can validate both leak tightness and functional performance (e.g., verifying that an orifice, restrictor, or internal channel is not blocked).

Important distinction: Flow tests can be confounded by internal geometry variation (especially in AM), surface roughness, or partially obstructed passages. Specify whether you are measuring external leakage, internal flow performance, or both.

5) Vacuum box / localized vacuum testing

This method uses a sealed vacuum box over a weld seam or surface feature, often with a bubble solution. It is effective for surface-breaking leaks on accessible areas but does not evaluate internal-to-external leakage pathways hidden within complex passage networks.

6) “Proof” testing vs. leak testing

Programs sometimes conflate pressure proof/burst testing with leak testing. A proof test demonstrates structural integrity at pressure (often at a multiple of operating pressure) but can still pass while a small leak exists. Conversely, an aggressive leak test pressure can damage thin-wall parts or temporarily seal a leak via deformation. Treat these as distinct requirements with separate acceptance criteria.

Pass/fail criteria

A leak test is only as meaningful as the acceptance criteria. The most common procurement failure is asking for “leak test per standard” without specifying the leak rate limit, test pressure, and test configuration. For manifolds, housings, and ducts, define pass/fail in engineering terms that tie directly to system risk.

Define the leak rate in the right units

Helium mass spectrometer testing typically reports leak rate in atm-cc/sec (also seen as std-cc/sec). Pressure decay often reports sccm, Pa·m³/s, or an equivalent derived leak rate. Avoid ambiguous language like “no leak” unless the method is explicitly qualitative.

Specify test pressure(s) and directionality

Leak paths can behave differently depending on pressure direction and magnitude (e.g., seals, thin walls, brazed joints). State:

• Operating pressure and required test margin (e.g., 1.1× or 1.5× operating) for leak testing
• Proof pressure if needed, and whether proof occurs before or after leak test
• Differential pressure across the boundary being tested (especially for assemblies with internal cavities)

Control temperature and stabilization time

Temperature drift can mimic leakage in pressure decay tests and can shift baseline in helium tests. Require the supplier to document:

• Stabilization time after pressurization/evacuation
• Test duration
• Temperature measurement location (ambient and/or part temperature)
• Compensation method (if used)

Define what constitutes the “test boundary”

Manifolds and housings often have multiple ports, threaded features, and machined sealing faces. Clarify what is included in the leak-tight boundary:

• As-machined ports (threads, O-ring glands, boss faces)
• Welds/brazes
• Inserted fittings (if part of delivered configuration)
• Sealed plugs/temporary closures (test tooling vs. product configuration)

Acceptance criteria should match risk and downstream integration

For example, a duct that only sees low-pressure airflow may tolerate a higher leak rate than an electronics housing requiring environmental sealing, or a manifold feeding a controlled pneumatic or propellant-adjacent function. The key is to translate system-level requirements into a numeric criterion that the supplier can test and report.

Preparing parts for testing

Leak testing is not a “plug it in and pressurize” activity for precision parts—especially when using AM + HIP + machining workflows. Preparation steps determine whether the test is valid and repeatable, and they often dictate whether you can localize a leak for corrective action.

Step 1: Confirm the manufacturing state (and freeze it)

Ask whether the leak test is performed as-built, after HIP, after heat treat, after CNC machining, or after coating/impregnation. For PBF DMLS/SLM parts, this sequence matters:

Typical critical workflow: build → stress relief → support removal → HIP (if specified) → heat treat (if required) → 5-axis CNC machining of sealing features → NDE as required (CT scanning, penetrant, etc.) → cleaning → leak test → final inspection (CMM dimensional + visual) → CoC / certification pack.

Why it matters: HIP can close internal porosity, but machining can expose near-surface pores; conversely, some coatings can temporarily mask leakage. Define the leak test at the state that best reflects delivered hardware and in-service condition.

Step 2: Cleanliness and drying

Residual machining coolant, oil, or cleaning fluid can block small leak paths or outgas during vacuum helium testing. Require a documented cleaning process appropriate to the material and application (e.g., stainless, Inconel, titanium, aluminum). For vacuum methods, specify that parts must be dry and free of volatile residues prior to test.

Step 3: Fixturing and port closure strategy

Most manifolds and housings require temporary closures to isolate cavities and create a defined test boundary. The closure method can introduce false failures (leaking test plugs) or false passes (overly compliant seals that mask poor sealing surfaces).

Best practice questions to ask:
• What fittings or plugs are used? (threaded, face-seal, O-ring, compression)
• Are sealing surfaces protected? (avoid galling of soft alloys or titanium)
• Are torque values controlled and recorded?
• How is test volume determined? (for pressure decay calculations)

Step 4: Masking “non-product” leak paths

Some components have features that are intentionally open during manufacturing (vent holes, temporary build drains, EDM start holes) that will be closed later. Ensure engineering explicitly identifies which features are to be sealed prior to leak testing and how (weld, braze, plug, fastener, sealant). The test plan should match the delivered configuration.

Step 5: Material and process considerations (AM, HIP, PM-HIP)

For AM PBF parts, leaks often originate from lack-of-fusion defects, near-surface porosity, or microcracks at stress concentrators. HIP can reduce internal porosity, but it does not automatically guarantee leak tightness if defects are connected to the surface or if machining opens a connected pore network.

For PM-HIP parts, densification is typically excellent, but leak risk can still come from:

• Canister-related issues (if applicable) or post-canister removal surfaces
• Machining-induced tearing at sealing faces
• Assembly joints added after densification (welds, brazes, inserts)

Align leak testing with the actual failure modes for the process route used.

Documentation

In regulated manufacturing, the test result is not complete without traceable evidence. Your leak test documentation should support internal quality review, customer audits, and downstream acceptance (including government/prime contractor flows). At a minimum, request a report that is unambiguous, calibrated, and tied to serial numbers.

What a good leak test report includes

• Part number and revision, serial number(s), and quantity tested
• Test method (helium mass spectrometer vacuum/pressure, pressure decay, bubble, etc.) and a brief setup description
• Acceptance criteria (leak rate limit, pressure, duration, stabilization time, allowable background level)
• Actual test conditions: test pressure/vacuum level, gas type and concentration (e.g., helium percentage), temperature/ambient conditions, time at pressure
• Measured results: numeric leak rate (or stated qualitative outcome, if applicable), plus background and final reading where relevant
• Equipment identification: instrument model/ID, fixture ID, and software revision (if automated)
• Calibration status: calibration due date and reference standards used (traceable calibration is key for AS9100 environments)
• Operator and inspector sign-off, date/time stamps, and revision-controlled procedure number

Integration with the certification pack

For aerospace/defense procurements, leak test data often sits inside a broader pack, such as:

• Certificate of Conformance (CoC)
• Material traceability: heat/lot, mill certs, powder lot (for AM), and build ID where applicable
• Process certifications: HIP records, heat treat charts, welding/brazing records, NDE reports (CT scanning, penetrant, etc.)
• Dimensional inspection: CMM reports for critical features and sealing surfaces
• First Article Inspection (FAI) documentation when required by contract/customer flow-down

If the part is ITAR-controlled or DFARS-sensitive, also ensure the supplier can document controlled handling, segregation, and record retention per contract requirements.

Typical pitfalls

Most leak test disputes trace back to preventable issues in the RFQ or the supplier’s test plan. These are the most common pitfalls seen on manifolds, housings, and ducts.

Pitfall 1: Calling out “helium leak test” without specifying vacuum vs. sniffer

Vacuum helium testing can detect much smaller leaks than a sniffer method. If you need low leak rates, specify vacuum mode (or specify the required sensitivity and let the supplier propose the configuration).

Pitfall 2: Not accounting for internal volume and temperature (pressure decay)

Pressure decay can be highly repeatable when engineered correctly, but it can also generate noise and false rejects if the internal volume is large/complex (common in ducts and manifolds) or if the part temperature changes during the test. Require stabilization time, temperature control/compensation, and a defined method for volume estimation.

Pitfall 3: Leaks in test tooling mistaken for part leaks

Threaded plugs, temporary fittings, and O-rings are frequent culprits. Demand that the supplier perform a tooling leak check (and document it) before testing production hardware, especially when using multiple closures on complex manifolds.

Pitfall 4: Inappropriate test pressure that damages thin-wall AM parts

DMLS/SLM ducts and housings may be thin and highly optimized. Excessive pressure can cause yielding, distortion, or crack growth, turning a screening test into a destructive event. Confirm allowable test pressure relative to design allowables and whether a proof test is required separately.

Pitfall 5: Testing too early (before final machining of sealing features)

If critical sealing faces, O-ring glands, or threaded ports are not finished, the test may be meaningless. Conversely, machining after leak test can open porosity and create new leak paths. Align leak testing with the final state of the sealing boundary.

Pitfall 6: Confusing external leakage with internal cross-leak

Manifolds often require both: (1) no leakage to atmosphere and (2) no leakage between isolated internal circuits. Specify whether you need a cross-port isolation test and how it is performed (which ports sealed, which pressurized, what constitutes a pass).

Pitfall 7: “No bubbles” acceptance criteria for critical hardware

Bubble testing is qualitative and operator-dependent. If the risk profile demands a numeric leak rate, specify a quantitative method and a leak rate limit.

RFQ language

Below is procurement-ready language you can adapt. The goal is to reduce interpretation, ensure comparable supplier quotes, and drive a documented, auditable result.

1) Define the test objective and boundary
Supplier shall perform leak testing on [PART NUMBER / REV] to verify pressure boundary integrity for [to-atmosphere leakage and/or internal cross-leak between circuits]. Leak test boundary includes all machined ports, sealing faces, and joints in the delivered configuration. Temporary closures used for testing shall not damage sealing surfaces.

2) Specify the method and sensitivity
Method: Helium mass spectrometer leak test in vacuum mode (preferred for high sensitivity), or supplier-proposed equivalent method meeting the acceptance criteria below. Supplier shall identify configuration (vacuum chamber vs. vacuum fixture), helium concentration, and stabilization time in the test plan.
Alternate: Pressure decay may be proposed for production screening if supplier demonstrates correlation to helium results and controls temperature/volume effects.

3) Provide numeric acceptance criteria (edit values to match your system requirement)
Acceptance: Maximum allowable leak rate = [X] atm-cc/sec helium equivalent at [TEST PRESSURE] and [TEMPERATURE RANGE]. Test duration shall be ≥ [TIME] after stabilization. Background helium level shall be recorded and controlled to prevent false positives.

4) Define pressures, sequences, and part condition
Part condition: Leak test shall be performed after HIP (if specified), heat treat, and final CNC machining of all sealing features, and after final cleaning and drying. If any coating, impregnation, or sealant is used, supplier shall disclose and obtain written approval prior to processing.
Sequence: If proof pressure testing is required, perform proof test at [PROOF PRESSURE] prior to leak test unless otherwise specified. Supplier shall report any permanent deformation, cracking, or functional change observed.

5) Require documentation suitable for AS9100 programs
Deliverables: Supplier shall provide a leak test report containing part identification (PN/Rev, serial numbers), method and setup, acceptance criteria, actual test conditions, measured results, equipment ID, and calibration status. Report shall be tied to the shipment CoC and material/process traceability records (including powder lot/build ID for AM parts, HIP records, and heat treat charts where applicable).

6) Address controlled handling (ITAR/DFARS) when applicable
Compliance: Supplier shall maintain ITAR-controlled handling and record retention per contract. If DFARS specialty metals or other flow-down requirements apply, supplier shall maintain traceability and provide objective evidence within the certification pack.

7) Ask for a pre-test plan and a failure response
Test plan: Supplier shall submit a test plan for approval prior to first article, including fixturing sketches, port closure method, stabilization times, and calculation method (if pressure decay).
Failure response: In the event of failure, supplier shall document suspected leak location, troubleshooting steps (e.g., sniffer localization), and disposition recommendation (rework, repair, or scrap) with engineering rationale.

Well-specified leak testing reduces late surprises and improves supplier accountability. If you define method, criteria, part condition, and documentation up front, you will get results that are comparable across suppliers, defensible in audits, and aligned with the real risks of flight- and mission-critical hardware.