Leak Test Specification for Pressure-Tight Aluminum Castings: Helium vs Pressure Decay, Acceptance, and Data

# Leak Test Specification for Pressure-Tight Aluminum Castings: Helium vs Pressure Decay, Acceptance, and Data

Many RFQs say "leak tight" and stop there. Suppliers then assume different test methods, different pressures, different acceptance thresholds, and different sampling plans. The quote becomes cheap on paper and expensive in launch.

This guide shows how to write a leak-test requirement that is quoteable and auditable for pressure-tight aluminum castings such as pump housings, valve bodies, hydraulic manifolds, compressor housings, coolant passages, and sealing-critical covers.

Helpful Bohua routes:

• •Leak-tight casting buyer route
• •Pump housing intent: leak risk + inspection inputs
• •Valve body intent: ports + sealing + inspection inputs
• •Request quote + drawing upload

1) Define what "leak" means for your program

Include, at minimum:

• •the medium (air, nitrogen, helium, water, oil, coolant, fuel, etc.)
• •operating pressure range and temperature range
• •the functional consequence of leakage (performance loss, safety risk, warranty risk, environmental exposure)
• •whether the leak path can be internal-only, external-only, or both

If you have a field standard or internal spec, attach it. If you do not, state that you need a supplier recommendation and want their assumptions written into the quote.

2) Choose a test method (and do not mix methods silently)

Common production methods include:

• •Helium leak test (very sensitive, often used for sealing-critical programs)
• •Pressure decay (air/nitrogen) test (practical for higher-volume lines)
• •Water immersion bubble test (useful but typically less quantitative)

Do not assume "leak test" equals helium. State the method you require, or explicitly allow alternatives if the supplier can justify equivalency.

3) State the test pressure and hold time

Two suppliers can both say "passed leak test" while using different pressures and different hold times.

Specify:

• •test pressure (absolute or gauge)
• •ramp profile if relevant (fast pressurization can mask or create false rejects)
• •stabilization time (thermal and fixture stabilization matters)
• •hold time (the measured window)

If your working pressure is not final yet, give a range and request a supplier recommendation for test pressure and hold time for each method.

4) Provide an acceptance threshold (leak rate or pressure drop)

Pick one measurable acceptance statement:

• •allowable leak rate (for helium or mass-flow)
• •allowable pressure drop over a defined time window for a defined volume

If you do not know the threshold, request the supplier to propose a threshold that matches your risk and the part function, and require them to document it in the quote and first-article records.

5) Define sampling vs 100% testing (and what changes the rule)

State:

• •whether you require 100% testing, sampling, or phased ramp-up (100% at launch, sampling after capability)
• •whether any repair or impregnation is allowed, and how it is recorded
• •whether any secondary test is required after machining or cleaning

If you are unsure, ask for two quote scenarios: 100% testing and sampling-based testing, both with the same acceptance threshold.

6) Clarify the "test stage" in the manufacturing flow

Leak results depend on when the test happens:

• •as-cast (before machining): useful for process health, but may not represent final sealing surfaces
• •after rough machining: exposes sub-surface porosity risk in machined ports and faces
• •after finish machining and cleaning: closest to end-use seal condition

Tell the supplier which stage you care about, or require both a process-health test and a final acceptance test.

7) Fixture, sealing, and interface assumptions must be written down

Leak disputes are often fixture disputes. Include:

• •which ports are sealed and how
• •whether plugs or o-rings are buyer-provided or supplier-provided
• •whether seal faces are machined or as-cast
• •whether any threads are part of the sealing interface

If the final interface is a mating assembly, attach the interface drawing or a simplified seal schematic.

8) Data package: what proof you expect per lot

Define what you want to receive:

• •lot-level summary (pass rate, reject reason categories)
• •test pressure, hold time, and acceptance threshold printed on the record
• •traceability fields (lot, heat, tooling version, date, shift)
• •whether you need per-part records, per-lot records, or only on request

If you need PPAP-style documentation, state that early so test records are designed with the same discipline.

9) What to ask suppliers to avoid "cheap quote, expensive launch"

Ask every supplier:

• •Which test method are you quoting, and why is it appropriate?
• •What are your quoted assumptions for pressure, hold time, and acceptance?
• •At which manufacturing stage is the acceptance test performed?
• •If a part fails, what is the allowed disposition (scrap, re-test, repair, impregnation)?
• •What evidence will you provide with each shipment (and what is stored for traceability)?

Start a structured RFQ (with leak-test prompts prefilled)

If you want the quote to include the leak-test assumptions in writing, use a structured RFQ entry so the requirement is visible to engineering from the first reply:

• •Start a leak-test focused RFQ
• •Or use the formal RFQ hub