Pneumatic Leak Testing Fixtures

A pneumatic leak-test fixture is part of the measurement system. Its seals, enclosed volume, stiffness, thermal behaviour, valves and connections can change the result as much as the instrument settings. This guide explains how to specify and qualify that system so a pass or fail decision describes the product under a defined test condition.

The page owns test volume, stabilisation, reference conditions and measurement capability. It includes six worked examples and the complete Leak test fixture requirement matrix. It does not set a universal leakage limit or test pressure. Product acceptance belongs to the responsible product authority, and safe test arrangements require competent design. The content is ready for technical and editorial staging review; it is not an approved procedure for an unidentified product.

1 The decision a fixture must support

Begin with the product decision. State what leakage matters, under which conditions, through which boundary and for what consequence. A production check for assembly defects may differ from a development measurement or an end-use performance requirement. The fixture and test method must be selected around that decision rather than around a convenient available instrument.

The words “leak free” are usually too imprecise for a specification. A measurement has a method, operating conditions, resolution, uncertainty and a decision rule. A result below a stated threshold is not proof that no molecule can escape under every possible condition. Write the acceptance requirement in measurable terms and retain the limitation in the report.

The fixture designer needs to know whether the test must detect an external leak, communication between internal cavities, a missing feature, a blocked passage or several defect types. A single pressure-decay result may not distinguish them. Some defects can produce an apparently good result if they prevent the intended volume from being tested.

2 Direct answer for an engineering enquiry

Specify a pneumatic leak-test fixture by the product boundary, authorised test pressure, allowable leakage with declared reference conditions, total effective test volume, sealing arrangement, stabilisation behaviour and demonstrated measurement capability. Include gross-leak and blocked-path challenges, fixture-background checks, calibration or verification requirements, traceable test records and a response to invalid results.

Compressor pressure and a sensitive leak tester do not by themselves establish capability. The station must prove that it tests the correct volume, reaches the intended condition, distinguishes relevant good and defective parts, and remains repeatable across the production conditions for which it is released.

3 Distinguish product leak testing from plant leak surveys

A plant leak survey seeks losses from a distribution system and supports repair prioritisation. A product leak-test fixture makes a decision about an individual assembly or defined sample under controlled conditions. Their instruments, acceptance criteria and records may differ even though both concern escaping air.

Their actual bodies and routes require reconciliation before use.

This page preserves that application-specific air-quality principle and develops the fixture measurement consequences. It does not transfer general shop-air assumptions into a precision test requirement.

4 Write the product requirement before choosing a method

The product authority should identify the relevant service condition and explain how the proposed production test relates to it. The test gas, direction of pressure difference, temperature, product state and leakage pathway all matter. A convenient air test cannot automatically prove liquid tightness, vacuum performance or behaviour at another pressure.

If a correlation between production test and service performance is required, define the evidence needed to establish it. Representative defects and product variation may be more informative than a single ideal reference or a theoretical conversion. Keep the correlation's range and exclusions in the validation record.

The requirement should also explain what happens to an invalid or inconclusive result. Treating every completed cycle as either pass or fail can hide a missing pressure condition, an instrument fault or an untested cavity. A separate invalid state protects the meaning of the pass result.

5 Define the test boundary on a drawing

Show the device under test, fixture seals, supply isolation, measurement connection and every volume exposed during measurement. Identify any internal cavities that should communicate and those that must remain separated. The drawing should show the actual test state, not only a general pneumatic circuit.

A cavity isolated by a product valve may remain untested even when the instrument reaches the programmed pressure. A fixture can also inadvertently seal over a defect or block a passage that should be exposed. The design review must therefore consider how the fixture changes the product boundary.

Use a controlled drawing for each product family or fixture variant. If an interchangeable insert changes volume or sealing position, that change belongs in the recipe and validation record. Similar external appearance is not evidence that two variants have equivalent measurement behaviour.

6 Select the method around the measurement problem

Pressure-decay testing observes a pressure change in an isolated volume under defined conditions. Differential pressure-decay testing compares the test side with a reference side and can reduce the influence of common effects when the arrangement is appropriately matched. Mass-flow methods measure the flow needed under a controlled condition. Each method has assumptions and limitations.

Manufacturer guidance from Furness Controls describes the importance of thermal and volume effects in pressure-decay measurement and the role of a reference side in differential arrangements [L1]. That general principle does not establish that any particular reference volume will cancel the thermal behaviour of the product.

The choice should consider product volume, compliance, thermal history, leak limit, test pressure, gas path and production constraints. Do not claim one method is universally most accurate or quickest. Ask the instrument supplier to substantiate capability for the actual fixture and requirement.

7 Distinguish pressure containment from leak acceptance

A leak test is not automatically a strength or proof test. Product pressure capability, fixture containment and personnel protection must be established independently. The required leak sensitivity cannot justify exposing an unqualified product or fixture to a higher pressure.

The test pressure may also change the defect itself. A seal can seat differently, a flexible wall can move, or a leakage path can respond nonlinearly. Testing at a higher pressure is not a universal way to accelerate an equivalent measurement.

Keep the authorised pressure envelope, product acceptance condition and instrument operating range in separate fields. This page identifies that interface without giving an operational pressure-testing procedure.

8 Account for all effective test volume

The relevant measurement volume includes the product cavity and connected fixture passages, tubing, valves and instrument spaces that participate in the measurement. A short piece of tubing can be significant for a small product. Conversely, the product may dominate in a large assembly.

Dead volume reduces the pressure change associated with a given mass loss under otherwise identical rigid and isothermal conditions. It can also lengthen filling and thermal settling. Reducing unnecessary volume can improve the measurement arrangement, but changes must preserve flow access to the intended cavity and avoid creating restrictions or trapped regions.

The geometric volume may differ from the effective volume used by an instrument's calibration method. Record how volume was established and whether compliance or calibration accounts for additional behaviour. Do not mix a geometrically calculated value with an empirically calibrated recipe without understanding the instrument's method.

9 Treat fixture compliance as a measurement input

Flexible tubing, elastomer seals and product walls can change volume as pressure changes or as materials relax after loading. Such movement changes pressure even without gas leaving the system. A rigid-volume conversion can then misrepresent apparent leakage.

Clamping can also affect the product. Excessive or poorly located force may temporarily close a defect, distort a seal or change the cavity volume. A fixture that produces repeatable numbers by deforming the product into an unrepresentative state may be consistently wrong for the intended decision.

Record the approved clamping arrangement, insert condition and product support. Where material relaxation is relevant, evaluate it over the intended cycle and product temperature range. A longer delay is not automatically the best remedy; the fixture or method may need redesign.

10 Design seals for both integrity and diagnostic access

Fixture seals need a defined contact surface, material, compression arrangement and replacement basis suitable for the product. They must contain the test condition without masking the leakage path being assessed. Sealing against a surface that varies significantly between parts can introduce fixture leakage that resembles product variation.

Provide a way to distinguish fixture background from product leakage. A suitable master or blanking arrangement can help, but it must represent the relevant seal and volume behaviour. An instrument connected directly to a blank is not a complete check of a fixture with several seals and passages.

Maintain the as-found condition when diagnosing a leak. Replacing several seals and changing clamping settings simultaneously may restore performance but leave the cause uncertain. Record the intervention and the requalification required before production decisions resume.

11 Specify air quality for the entire test station

The instrument and fixture require air suitable for their materials, valves, sensors and measurement method. Moisture, particles or oil can affect seal behaviour, contaminate references or restrict small paths. The applicable instrument manual and product cleanliness requirement should determine treatment; there is no universal quality class for every fixture.

Separate the supply for test measurement from other station uses where the design requires it. Clamping actuators or blow-off can create pressure disturbances or contamination pathways. A common upstream source may be acceptable, but the branch arrangement must demonstrate the required stability and quality at the instrument inlet.

A clean-air claim needs a location and evidence. The compressor outlet, the final filter outlet and the fixture cavity are different points. Retain the supplier's stated treatment conditions and the maintenance practices required to preserve them.

12 Understand what stabilisation is intended to remove

Filling can change gas temperature. The product and fixture can exchange heat with the gas, and mechanical elements can relax. During the measurement interval, these effects can appear as pressure loss or gain even when the gas mass is unchanged.

Stabilisation should be established from observed behaviour under representative conditions. It is not a universal number of seconds. A cycle validated after a long idle period may behave differently during repeated production, after a product-temperature change or following a fixture insert replacement.

Longer waiting may reduce some thermal effects while increasing cycle time and exposure to other drift. Select the interval from evidence of measurement capability and the product requirement. Do not tune the timer only until a preferred part begins to pass.

13 Keep pressure temperature and volume effects separate

For an ideal gas, pressure is related to gas amount, absolute temperature and volume. A falling pressure may result from gas loss, cooling, volume growth or a combination. A rising temperature can partly conceal a true leak, creating a false pass if the interpretation assumes constant temperature.

A temperature sensor on the fixture surface may not represent the mean gas temperature during a short transient. Compensation needs a validated relationship to the actual measurement, not just the availability of a temperature input.

Differential measurement can reduce effects shared by test and reference sides, but unequal materials, volumes, thermal histories or restrictions can leave a residual difference. Validate the assembled arrangement over its intended range rather than describing differential measurement as immune to temperature.

14 Understand the pressure-decay conversion

Under the simplified assumptions of fixed rigid volume, constant uniform temperature and ideal-gas behaviour, the equivalent reference leak flow is the test volume divided by reference absolute pressure, multiplied by the ratio of reference temperature to test temperature, multiplied by the rate of pressure decrease.

The pressure decrease is expressed in pressure units per time. The reference pressure is absolute. Use consistent units throughout, then convert the resulting volume-per-time to the declared reporting unit. A gauge-pressure change can represent the same absolute-pressure change only where the reference atmospheric pressure is effectively unchanged over the interval and the measurement arrangement supports that interpretation.

This relationship describes an average equivalent flow over the chosen interval. It does not account automatically for a leak path whose flow changes with pressure, thermal drift, flexible volume or a replenishing supply leak. The instrument's actual algorithm and calibration method must be understood before using the equation as a check.

15 Define leakage units completely

A reference volumetric flow needs the reference pressure and temperature and, where relevant, the gas and humidity basis. “Sccm” is often used for standard cubic centimetres per minute, but the word standard does not identify one universally shared temperature. Record the actual convention.

Pressure-volume throughput units, such as mbar L/s, are not numerically interchangeable with a reference volumetric flow. A conversion requires the relevant reference pressure and compatible temperature and gas assumptions. A result without its basis can appear different simply because it uses another convention.

A pressure drop in Pa or mbar is also not itself a leak flow. It becomes a flow estimate only through a model or calibration that accounts for the effective volume and timing. The same pressure drop from two different volumes does not necessarily represent the same gas loss.

16 Establish a valid reference and verification strategy

A known low-leak master, a calibrated leak and a fixture blank serve different purposes. The master can represent product behaviour; the calibrated leak introduces a known flow under specified conditions; the blank can isolate parts of the fixture background. They are not interchangeable simply because each is called a standard.

Cincinnati Test Systems describes a pressure-decay conversion method using a known non-leaking part and a leak standard, with a repeatable test sequence [L2]. That is evidence for the need to understand the instrument's calibration approach. It is not a universal instruction to apply one manufacturer's sequence to every tester.

A calibrated leak's stated flow depends on its specified conditions. Keep its identity, certificate, pressure, gas, direction and reference basis with the recipe. Inspect and handle it according to its documentation; contamination or an unintended connection arrangement can invalidate the assumed challenge.

17 Separate calibration from application validation

Instrument calibration establishes characteristics of the instrument or reference under its stated conditions. Application validation demonstrates that the assembled station makes the required product decision. A current calibration label cannot prove the fixture reaches every intended cavity or detects the relevant production defects.

Validation should include representative acceptable parts, relevant defective conditions and credible invalid states. It should cover the product variants, thermal conditions, fixtures, operators or loading arrangements that can materially affect the result. The scope must be justified by the application rather than copied from a generic sample-count rule.

Keep calibration status and validation status as separate records. After a change, one may remain valid while the other needs review. Replacing a fixture insert may not alter an instrument's calibration but may change volume, sealing and thermal response enough to require application requalification.

18 Make gross leaks and blocked paths visible

A fine-leak measurement needs the intended test condition to exist. A gross leak may prevent pressure from being reached. A blocked connection may allow a small upstream volume to reach pressure while the product remains untested. Both require explicit detection logic or another qualified verification method.

Challenge the station with conditions that represent these failure modes safely within the approved validation arrangement. Confirm the result is an appropriate fail or invalid state and that the record cannot be mistaken for a valid pass. A fast pressure rise is not automatically proof of a well-sealed product; it may indicate unexpectedly small connected volume.

A missing part can also produce misleading behaviour if fixture seals close against each other or a valve isolates the instrument. Presence detection, recipe control and pneumatic evidence may each contribute, but their actual effectiveness must be demonstrated.

19 Identify replenishment paths that can hide leakage

A pressure-decay measurement assumes the test volume is isolated from the supply during the measurement interval. If a valve leaks inward from a higher-pressure source, it can offset gas escaping through the product. The displayed decay may then understate the true product leak.

The circuit review should identify all potential replenishment paths, including shared manifolds, test-port connections and cross-cavity arrangements. Verification should follow the instrument and fixture design, with safe methods appropriate to the equipment.

Do not conclude that a station is healthy only because a leak-tight master gives a low decay. A replenishing path could make that observation look reassuring while reducing sensitivity to a defective part. A relevant known-leak challenge provides different evidence.

20 Evaluate repeatability reproducibility and drift

Repeatability concerns variation under substantially the same measurement conditions. Reproducibility examines relevant changed conditions, such as operator, fixture nest or loading method. Drift concerns change over time. Each can affect the reliability of a pass/fail decision near the acceptance limit.

Use a study design that reflects the production decision. Repeating one part without removing it may omit loading and seal variation. Repeatedly testing a part can also change its temperature or mechanical state, so the study itself needs controlled interpretation.

Report distributions and conditions, not only an average. A small mean error can coexist with large variation. Avoid a universal percentage-of-tolerance rule unless the product authority has selected and justified it for the risk and measurement context.

21 Set the decision rule before seeing the production result

The acceptance limit and the rule for accounting for measurement uncertainty should be agreed before routine testing. Depending on the application, a measured value near the limit may require a guard band, an inconclusive category or another defined disposition. The product authority must decide the acceptable decision risk.

Do not repeatedly retest a failed part until it produces a pass without an authorised retest rule. Thermal changes, reseating or chance variation can shift the result. Selective retention of passing cycles destroys the traceability of the original decision.

The system should retain the first result, reason for retest, intervention if any, subsequent results and final disposition. A retest policy is part of product quality control, not a hidden operator workaround.

22 Estimate air demand without confusing it with leak sensitivity

Station consumption includes charging the test volume, fixture actuators, purges or other approved uses and any instrument-specific flow. The average demand depends on cycle frequency, but the branch must also accommodate the filling transient without violating the instrument's supply requirements.

A low leakage acceptance limit does not imply negligible station air demand. Most consumption may occur during filling and exhaust rather than through the measured product leak. Conversely, increasing available compressor flow does not improve the instrument's leak resolution by itself.

Supply assessment should use the approved cycle and instrument requirements. Where several stations share a branch, account for overlapping fills and other users on a stated time basis. Do not multiply a guessed peak by a universal diversity factor.

23 Manage recipes and product variants

A recipe should identify product and fixture revision, method, pressure basis, timing, volume or calibration basis, limits, decision rule and required reference checks. Control access to changes that can affect product disposition. The station should make the active recipe and part identity visible in its records.

Do not assume one recipe covers every part with the same connection. Different cavity volumes, materials, seals or internal configurations can change response. Where a family recipe is justified, document the envelope and the evidence that supports it.

After a software update or recipe transfer, verify more than the displayed name. Units, decimal precision, hidden settings, result mapping and communications can change the meaning of a result. Preserve a recoverable version of the released configuration.

24 Worked case one on converting a pressure drop

Illustrative calculation only. Assume a rigid total test volume of 2.000 L, a pressure decrease of 100 Pa during 10.0 seconds, constant gas temperature of 293.15 K and reference conditions of 101,325 Pa absolute at 293.15 K. Treat the gas as ideal and assume no replenishment or thermal effect.

The pressure-decrease rate is 10.0 Pa/s. Equivalent reference flow is 0.002000 × 10.0 ÷ 101325 = approximately 0.0000001974 m³/s. Multiplying by one million to obtain mL/s and by 60 gives approximately 11.84 mL/min on the declared reference basis.

This number is an estimated average leakage under the assumptions. It is not a recommended acceptance limit and it does not establish instrument capability. The product authority must supply the limit and the station must demonstrate that non-leak effects are controlled adequately.

If the result is reported as sccm, the report must retain the stated reference temperature and pressure. Otherwise another reader may compare it with a value based on a different standard convention.

25 Worked case two on omitted fixture volume

Assume the product cavity is 1.000 L but the connected fixture, tubing and instrument spaces add 0.500 L. The relevant rigid volume is therefore 1.500 L. For the same pressure-decay rate, using only the product volume would calculate two-thirds of the flow estimated using the full volume.

If the actual equivalent flow were illustratively 12.0 mL/min under the simplified model, the incomplete 1.000 L calculation would report 8.0 mL/min. The result is 33.3 percent below the full-volume value. This is a volume-accounting error, not improved sealing.

The correction is to establish the appropriate effective volume or calibration basis for the actual instrument and fixture. Do not simply add a geometric volume to a recipe that already incorporates volume empirically; that could double-count the correction.

The example also explains why shortening a hose after validation is a controlled change. It may be beneficial, but it changes the relationship between pressure decay and leakage and can alter fill and settling behaviour.

26 Worked case three on a small thermal change

Assume a closed rigid gas volume with no leakage at an absolute pressure of 300,000 Pa and temperature of 300.00 K. If the gas cools by 0.10 K while volume and gas amount remain constant, the ideal-gas pressure change is approximately 300000 × 0.10 ÷ 300 = 100 Pa downward.

That pressure change is the same magnitude as the measured drop in case one, despite zero gas loss in this constructed example. It demonstrates why pressure decay cannot be interpreted as leakage without addressing thermal behaviour.

The example does not claim a surface sensor can measure or correct the gas temperature to that accuracy. It also does not prescribe a stabilisation time. The station must demonstrate performance under the actual thermal conditions and measurement method.

Heating can act in the other direction and mask some leakage. A validation that examines only false rejects from cooling may miss a false-pass mechanism under another thermal history.

27 Worked case four on pressure volume throughput

Assume a leakage throughput of 0.020 mbar L/s, with a compatible temperature basis for conversion. At a reference pressure of 1013.25 mbar, the equivalent reference volumetric flow is 0.020 ÷ 1013.25 = approximately 0.00001974 L/s.

Converting litres to millilitres and seconds to minutes gives approximately 1.184 mL/min on that reference basis. The numerical value 0.020 must not be entered directly into a field labelled mL/min or sccm.

The example is a unit reconciliation, not a leak-specification conversion between service conditions. If temperature bases differ, or if the stated throughput follows a different definition, the conversion must be revised. The original requirement and the converted value should remain linked in the record.

28 Worked case five on average charging demand

Assume an idealised 0.500 L test volume initially at 1.00 bar absolute and filled to 3.00 bar absolute at constant temperature. Use a reference pressure of 1.00 bar absolute at the same temperature. The net added reference gas volume is 0.500 × (3.00 − 1.00) ÷ 1.00 = 1.000 L per cycle.

At six cycles per minute, the idealised average charging demand is 6.000 reference L/min. This excludes actuators, purges, instrument consumption, leakage and any different exhaust or recovery arrangement.

If the fill occurs during only one second of each cycle, the mean addition during that one-second interval is 60 reference L/min. The instantaneous profile may differ. Thus an average of 6 L/min does not demonstrate adequate branch response during filling.

Neither pressure in this example is a recommended test condition. The purpose is to separate average compressor demand from the supply behaviour needed to maintain a valid test cycle.

29 Worked case six on an explicit decision rule

Assume the product authority sets an illustrative upper leakage limit of 10.0 mL/min and an illustrative expanded measurement uncertainty of 1.0 mL/min for the validated conditions. Suppose it chooses a conservative acceptance rule requiring the measured value plus that uncertainty to be no greater than the limit.

A measured 8.5 mL/min satisfies that particular acceptance rule because 8.5 + 1.0 = 9.5 mL/min. A measured 9.4 mL/min does not satisfy it because the corresponding value is 10.4 mL/min. That second result is not automatically proof that the true leakage exceeds 10.0; its disposition follows the agreed rule.

The uncertainty value and rule are assumptions for illustration. They are not recommended universal margins. The actual uncertainty evaluation must address the relevant measurement effects, and the product authority must define acceptance, rejection and any inconclusive category.

A display rounded to whole units could hide the distinction between these values. Reporting precision and stored raw results should support the approved decision rule without suggesting more accuracy than the system demonstrates.

30 Diagnose the station before blaming the product

A rising reject rate may reflect product variation, fixture wear, thermal conditions, supply disturbance or a configuration change. Preserve evidence before adjusting timers or limits. A change that reduces rejects is not necessarily a correction if it also reduces defect sensitivity.

Compare known references, product results, pressure traces, temperatures where relevant, fixture identity and recent interventions. Keep observations separate from suspected causes. If several changes are made together, acknowledge the limitation on root-cause attribution.

Use the following table to structure investigation under the approved maintenance and validation arrangements. It does not authorise work on pressurised components or bypassing guards.

31 Resolve fixture leakage without hiding it in an offset

A stable and understood background may be part of a validated measurement model, but an unexplained offset should not be subtracted simply to make results acceptable. Fixture leakage can change with seal wear, temperature, loading or product surface condition.

Establish whether the observed background comes from gas loss, thermal drift, compliance or the instrument's zero behaviour. These mechanisms do not necessarily respond to the same correction. A numerical offset can conceal the problem while leaving defect sensitivity uncertain.

After repair, verify the station using both a suitable low-leak reference and a relevant known-leak challenge. A low baseline alone does not show that the measurement path is intact or that the station can distinguish a part near the decision boundary.

32 Protect the validity of the data path

The value shown on the instrument, the result sent to the controller and the status stored against the part should agree. Units, scaling, sign convention, recipe number and invalid-result codes need explicit mapping. A technically sound pneumatic measurement can still produce incorrect product disposition through a data-handling error.

Test communications loss, duplicate part identifiers, interrupted cycles and recipe mismatch within the approved validation plan. The system should not reuse the previous pass result for a new part or treat an absent message as a pass. Define the permitted recovery and operator response.

Retain enough information to reconstruct a disputed result: part identity, fixture or nest, recipe revision, time, measurement value, units, status and relevant reference-check condition. Storage duration and access control follow the product and business requirements; this guide does not invent a universal retention period.

33 Balance throughput with capability

Cycle time is a design constraint, but it cannot be reduced independently of the physics. Filling, settling and measurement each serve a purpose. A shorter cycle must demonstrate that the required defect discrimination remains valid over the intended operating range.

Parallel fixtures may improve throughput while introducing differences between nests. Each relevant configuration needs evidence, and shared supply or temperature effects must be considered. Copying a successful recipe to another nest is a starting point for review, not proof of equivalence.

When the required throughput and capability cannot both be achieved with the proposed arrangement, make the conflict explicit. Options may include fixture redesign, reduced dead volume, improved thermal control, a different method or additional stations. The appropriate choice depends on validated results and commercial constraints.

34 Specify the station through a technical comparison

A useful enquiry gives suppliers the product boundary, intended defect detection, acceptance basis and representative conditions. It requests evidence of capability and identifies who supplies masters, calibrated leaks and validation samples. An instrument brochure alone is not a complete station proposal.

Compare proposed methods on the actual application. Include fixture design, supply requirements, cycle definition, maintenance, recipe control, data integration and verification.

35 Plan commissioning around the complete measurement chain

Commissioning should begin with the approved product requirement and fixture drawing. Confirm the installed configuration, measurement path, supply conditions, instrument status and recipe revision. The plan should identify who witnesses each acceptance item and where results are retained.

Demonstrate that acceptable and relevant defective conditions produce the intended decisions under the released conditions. Include invalid states such as failure to reach the test condition or a blocked path where those are credible. Do not report only a sequence of successful reference tests.

Where production extremes cannot be reproduced during commissioning, state how they are covered and what remains conditional. A laboratory demonstration at one temperature does not automatically validate the full factory environment or product mix.

36 Define release and requalification triggers

Release should identify the exact product and fixture revisions, method, recipe, operating envelope, decision rule and supporting evidence. It should also name the authority that accepts residual limitations. A supplier's installation sign-off and the product authority's measurement release are different records.

Requalification triggers may include changes to seals, tubing, valves, fixtures, software, calibration basis, product material or process temperature. The required scope depends on the change's effect. A documented assessment can justify a focused verification, while a major boundary or method change may require broader validation.

Keep superseded recipes and results traceable. Do not overwrite a prior configuration and then attempt to interpret old production results using the new recipe. Historical records need the conditions under which they were actually produced.

37 Required asset Leak test fixture requirement matrix

Complete this matrix before approving the station specification. Attach evidence or references where a short field cannot describe the requirement. Blank fields are information gaps, not permission to use default instrument settings.

Document identity: project ______; product family ______; product drawing and revision ______; fixture identification ______; matrix revision ______; preparer ______; product authority ______; fixture supplier ______; instrument supplier ______; date ______.

38 Required asset volume and measurement schedule

Measurement definition: reported quantity ______; pressure measurement type ______; reference pressure ______; reference temperature ______; gas ______; moisture convention if relevant ______; timing definition ______; conversion or calibration method ______; rounding and stored precision ______.

The schedule must distinguish an independently calculated volume from a value obtained through the instrument's calibration. Do not add both without a documented model. If volume cannot be represented adequately by one number, record the method that accounts for the actual behaviour.

39 Required asset validation challenge schedule

A challenge must be relevant and safely implemented. The matrix does not instruct anyone to create hazardous defects or defeat protection. The validation owner should approve representative samples and challenge methods within the qualified test arrangement.

40 Required asset commissioning and acceptance record

Record the installed fixture revision ______; instrument identity and calibration status ______; supply verification ______; active recipe and checksum or version reference ______; reference artefact status ______; environmental and product conditions ______; validation report ______.

Acceptance decision: released ______; conditionally released ______; not released ______. Scope of release ______; excluded variants or conditions ______; open items ______; owner for each item ______; closure evidence ______; approving roles and dates ______.

Attach the raw and summarised results needed to support the decision. If an average is reported, retain the observations and conditions behind it. If a result is excluded, record the reason and authority rather than silently removing it from the study.

Return-to-service record: change made ______; as-found condition ______; affected measurement assumptions ______; verification scope ______; reference results ______; recipe revision ______; release authority ______. This provides a reusable bridge between maintenance and product quality.

41 Required asset diagnostic and retest record

Record product identity ______; first result and status ______; recipe ______; fixture or nest ______; time and relevant conditions ______; observed issue ______; suspected cause ______; supporting evidence ______; authorised action ______.

Retest reason ______; intervention before retest ______; subsequent results ______; final disposition ______; approving authority ______; linked corrective action ______. Preserve all results required by the agreed traceability process.

A retest after reseating a part is not the same condition as a repeated measurement without unloading. A retest after a long delay may also involve a different thermal state. These distinctions can explain variation and should remain visible in the record.

42 Questions about leak test fixtures

Can a pressure decay result be converted directly into a leak rate

Only with a valid volume and condition basis or the instrument's established calibration method. Temperature change, compliance and replenishment can alter the result. The conversion must match the actual fixture and measurement interval.

Why does the first product after a break behave differently

The product, fixture and gas may have a different thermal or mechanical history after idle time. Supply and seal behaviour can also change. Investigate with reference checks and time records before changing acceptance limits.

Does a calibrated instrument prove the fixture is capable

No. Application validation must also show that the intended cavity is tested, relevant defects are detected and the assembled system performs across its released conditions.

Can fixture leakage simply be subtracted from every result

Only if a justified and validated measurement model supports the treatment. An unstable or unexplained background should be investigated rather than hidden by an offset.

Does differential pressure testing remove every temperature effect

No. It can reduce shared effects under a suitable arrangement, but mismatched thermal behaviour or other differences can remain. Capability must be demonstrated for the actual test and reference sides.

Why might a blocked product pass

The instrument may pressurise only a smaller upstream volume while the intended cavity remains isolated. The validation must include a relevant way to detect an incomplete test path.

Should a failed product be tested repeatedly until it passes

No. Retesting must follow an approved rule that preserves the first result, reason, intervention and final disposition. Selective acceptance of a later pass can conceal measurement or product problems.

What is the correct stabilisation time

There is no universal value. Establish it from the product, fixture, method and operating conditions, and verify the resulting capability over the intended production envelope.

Is a higher test pressure always more sensitive

No universal conclusion follows. Pressure can change leakage paths, product deformation and safety requirements. Use the product authority's approved condition and a validated method.

Can the same recipe be used on every fixture nest

Only when the evidence supports equivalence over the released conditions. Differences in volume, seals, temperature or calibration can require separate treatment.