How to Perform an Airtight Test on a Protective Case: The Pressure Decay Method in Practice

The short answer: the industry-standard way to check a protective case for airtightness is the pressure decay method. You pressurise (or evacuate) the sealed cavity to a set level, let the pressure stabilise, then measure how much the pressure changes over a defined interval. The result is not a subjective "it leaks or it does not" verdict but a quantified decay curve, which is exactly why the method suits 100 percent production inspection. To make the numbers trustworthy you must lock down five variables at the same time: test pressure, stabilisation time, measurement time, volume correction and temperature correction. Temperature drift is the most commonly ignored of the five and the most frequent cause of a false verdict.

Plenty of buyers and quality engineers accept a case on the strength of a single IP67 report, without asking what process capability sits behind that report. An IP rating is a type-test conclusion: it proves that the submitted sample did not admit water under specific conditions. It does not prove that every production unit has the same gasket compression, the same lid flatness, or the same latch clamping force. Airtight testing is the bridge that translates a type-test conclusion into a controllable production parameter. It screens every unit for leaks at a fraction of the cost and cycle time of an immersion test, and it turns an invisible sealing defect into a visible pressure number.

This article is written for procurement, quality, process and equipment engineers working with protective cases. It works through the physics of pressure decay, the standards that apply, how a test system is built, how the parameters should be derived, how temperature and volume corrections work, how the four families of leak test methods compare, how to localise the leak once you find one, and how to put the whole thing on a production line with traceable data. Every pressure, timing value and threshold quoted here is a typical or empirical figure. Binding acceptance criteria should always come from the product drawing, the customer technical agreement and the current text of the standard.

Contents

• What the Pressure Decay Method Actually Measures
• Why Protective Cases Need Airtight Testing: The Hidden Risk Beyond IP67
• The Physics and the Acceptance Formula
• The Relevant Standards: ASTM D4991, ASTM D3078 and IEC 60529
• Building a Test System: Air Supply, Fixtures, Sensors and Data
• The Procedure: Ten Steps from Preparation to Verdict
• Setting the Parameters: Pressure, Stabilisation, Measurement Time and Threshold
• Temperature and Volume Correction: Where Most Errors Come From
• Comparing the Four Families of Leak Test Methods
• Locating the Leak: Gaskets, Latches, Valves and Inserts
• How a Pressure Equalisation Valve Changes the Test
• Putting It on the Line: Full Inspection, Sampling and Traceability
• Common False Verdicts and a Troubleshooting Table
• Frequently Asked Questions
• Conclusion and Related Reading

What the Pressure Decay Method Actually Measures

Start with a one-sentence definition. The pressure decay method establishes a pressure differential across the wall of a sealed cavity, then observes how quickly that differential decays with time. If the case is perfectly sealed, the differential should hold essentially constant. If a path exists, gas migrates from the high-pressure side to the low-pressure side and the differential falls. The rate of decay relates directly to the equivalent flow area of the leak, the magnitude of the differential, the viscosity of the gas and the internal volume of the cavity.

Three concepts get mixed up constantly and need separating:

1. Airtightness. The inability of gas to cross the sealing interfaces of the case. It is a continuous quantity, expressible as a leak rate in millilitres per minute at standard conditions, or pascals per second.
2. Water resistance. The inability of liquid water to enter the enclosure under a given pressure for a given time. IPX7 is typically one metre of water for thirty minutes; IPX8 is by agreement at a greater depth or a longer duration.
3. The IP rating. The classification code defined in IEC 60529 and its Chinese counterpart, which labels the combined dust and water capability.

The relationship between the three is simple: airtightness is the mechanism, water resistance is the outcome, and the IP rating is the standardised label applied to that outcome. A case with poor airtightness can never reliably pass IPX7. The reverse does not hold: a sample that passed IPX7 is not proof that every production unit will pass an airtight test. That asymmetry is precisely why airtight testing occupies the role of a process control tool in this industry rather than a substitute for certification.

The most valuable output of airtight testing is not the word "pass". It is the distribution of leak rates. When the average decay value of a batch drifts from 0.3 to 0.9 in whatever units the equipment uses, the batch has not necessarily failed yet, but it is telling you that a gasket lot, an injection parameter or a latch clamping force has moved. You can intervene before an immersion test fails and before a customer complaint arrives.

Why Protective Cases Need Airtight Testing: The Hidden Risk Beyond IP67

Immersion testing is the most intuitive validation method and the least suitable for production screening. Three reasons.

• It is slow. The typical IPX7 condition is one metre of water for thirty minutes. Add loading, unloading, opening, inspection and drying and a single cycle usually runs past forty-five minutes, which no line producing a thousand units a day can absorb.
• Its verdict is soft. Judging by "open it and see whether water got in" invites error. Trace ingress can be absorbed by foam, masked by condensation, or leave a nearly invisible mark on contents.
• It is not reversible. A case that has been immersed must be dried thoroughly. Ship it damp and you invite mould or corrosion at the customer site.

Pressure decay compresses the cycle to tens of seconds or a few minutes, rests on a numeric decay value, and never wets the product. Its limitation is that airtightness and water resistance do not share a scale: the flow aperture that matters for gas is not the same as the paths available to liquid water. Pressure decay therefore belongs in the role of process control and screening; type validation still requires immersion testing to IEC 60529.

Combining the two is the sound approach. Validate the IP rating by immersion at the type-test stage. Once the design is frozen, take a unit that passed immersion and use it as a golden reference to calibrate the pressure decay threshold. Then screen production with pressure decay and re-verify the equipment periodically against the golden reference. The design logic behind that chain is set out in how an IP67 protective case is engineered, and the immersion procedure itself is covered in how the IP67 submersion test is run.

One further hidden risk deserves its own paragraph: seal drift after transport. A case that has been through vibration and impact in the logistics chain may end up with slightly relaxed latches, a marginally warped rim, and a changed gasket compression set. If airtight testing only happens before dispatch, the failure chain of "it only leaks after the journey" is never covered. The relevant validation method is described in vibration testing and transport resonance.

The Physics and the Acceptance Formula

Pressure decay rests on the ideal gas law. For a closed cavity of volume V at temperature T, the relationship between the amount of gas and the pressure is:

P V = n R T

When gas escapes, the molar quantity changes and the pressure follows. In practice the simplified engineering criterion compares the pressure drop across the measurement window against a threshold:

dP = P1 minus P2 is less than or equal to the allowed dP

Here P1 is the cavity pressure at the end of stabilisation, P2 is the pressure at the end of measurement, and the allowed value is the acceptance threshold. To convert the drop into an equivalent leak rate:

Q = (dP x V) / (P atm x dt)

Where Q is the equivalent volumetric leak rate referenced to standard conditions, V is the total cavity volume including fixture and manifold dead volume, P atm is local atmospheric pressure, and dt is the measurement time. Three consequences follow directly:

1. The larger the volume, the smaller the pressure drop produced by the same leak. Large cases therefore need longer measurement times or more sensitive sensors. A threshold lifted from a small case is meaningless on a large one.
2. The longer the measurement, the smaller the leak that can be detected. But the longer the window, the more temperature drift accumulates, so there is an optimum range rather than a monotonic benefit.
3. The threshold must be bound to the volume. If a supplier quotes a leak rate figure during negotiation, always ask for the corresponding cavity volume and test pressure. Without them the number carries no information.

In production, the more common implementation is not absolute leak rate but differential pressure decay. The part under test and a known-good reference part are connected to opposite sides of a differential pressure sensor and pressurised together. Because both sides share the same volume and the same thermal environment, slow ambient temperature changes act on both and cancel. The sensor reading then reflects only the difference in leakage between the two parts. This arrangement is far more immune to thermal drift than a single-chamber absolute pressure method and is the first choice for automated lines.

The Relevant Standards: ASTM D4991, ASTM D3078 and IEC 60529

There is no single standard that governs every airtight test. Different case types and different criteria point to different documents. Three are cited most often in engineering practice.

• ASTM D4991 covers leakage testing of empty rigid containers by the vacuum method. It applies to an empty rigid container, establishes a vacuum outside the container, and observes the container's internal pressure or an indicator to judge leakage or seal integrity. It suits a fully assembled empty case and is one of the closest ASTM methods to what the industry calls airtight testing of a protective case.
• ASTM D3078 covers the bubble emission method for leaks in flexible packaging. The specimen is immersed and continuous bubbles reveal leakage and its location. This is a bubble method rather than a decay method, but it is extremely useful for localisation and is often paired with pressure decay: screen with decay first, then pinpoint with bubbles.
• ASTM F2338 covers non-destructive leak testing of packages by vacuum decay. Its emphasis on non-destructive testing suits finished, packed product.

For the water resistance rating itself, the authoritative references are IEC 60529 and its Chinese equivalent. The two share essentially the same level structure and test conditions: IPX7 is short-term immersion, typically one metre for thirty minutes, while IPX8 is agreed between manufacturer and user and is normally deeper or longer. The IP6X dust level requires that no dust enter after exposure in a specified dust chamber.

One warning matters here. IEC 60529 does not define airtight test methods or thresholds. It defines how enclosure protection levels are classified and verified. A requirement phrased as "perform airtight testing in accordance with IEC 60529" is a category error. The correct phrasing is "verify the IP rating in accordance with IEC 60529, and screen production using pressure decay per ASTM D4991 or an agreed in-house method". Keeping those two activities distinct is the single most effective way to avoid disputes over technical documents. For the meaning of the rating itself see what the IP67 rating means, and for verification of the paperwork see how to verify an IP67 certificate.

Building a Test System: Air Supply, Fixtures, Sensors and Data

A workable pressure decay system has six subsystems, and a weakness in any of them shows up directly as untrustworthy data.

1. Air supply and regulation. Clean, dry compressed air or nitrogen is standard. Moist or oily air forms droplets in the manifold and produces drifting readings. A three-stage filter for particulates, water and oil, followed by a precision regulator holding output within about one percent of range, is the usual arrangement.
2. Fill and isolation valves. Closing speed and internal leak rate are the critical properties. A valve that leaks internally makes "the case leaks" and "the valve leaks" indistinguishable, and this is the most common source of false positives.
3. Pressure or differential sensors. Match the range to the test pressure rather than buying one large-range sensor for everything. If the test pressure is 20 kPa, a 0 to 100 kPa sensor usually resolves better than a 0 to 1 MPa unit. A resolution of about 0.1 percent of full scale is a reasonable target.
4. Fixtures and sealing tooling. The fixture temporarily seals any openings such as cable ports, valve ports or connector cutouts, and must not itself become a leak path. The sealing faces of the fixture should be clearly distinguishable from the product sealing faces; otherwise the leak cannot be attributed to the right component.
5. Temperature compensation. High-accuracy systems place a temperature sensor inside the cavity and correct the pressure reading in real time. Systems without compensation must rely on a long enough stabilisation period to reach thermal equilibrium.
6. Data acquisition and pass/fail software. Capture the pressure curve, measurement time, ambient temperature, equipment identifier and operator, and emit a traceable verdict. For supply into automotive, medical and defence channels, the traceability of the airtight test data is as important as the test itself.

The Procedure: Ten Steps from Preparation to Verdict

The following sequence applies to finished or semi-finished cases alike and can be adapted to the product.

1. Confirm the state of the part. Establish whether it is an empty shell, a foamed case or a finished pack, and whether the pressure valve, cable glands and inserts are already installed. Parts in different states must not share a single acceptance batch.
2. Seal every opening. Use dedicated tooling for cable ports, valve ports and any non-sealing interface. Each sealing element needs its own verification record.
3. Set the test pressure. Base it on wall thickness, case stiffness and customer requirements. Too high and a thin-walled case bulges elastically, which can itself disturb the seal; too low and the leak signal is too weak. Typical working values fall in the range of a few kilopascals to a few tens of kilopascals gauge.
4. Fill. Open the fill valve and pressurise the cavity. Fill time depends on cavity volume and supply flow: the larger the cavity, the longer the fill.
5. Stabilise. Close the fill valve and let the gas temperature equalise with the case wall. This is the stage most often cut short. If stabilisation is too brief, the heat generated by compression has not dissipated; the measurement window then shows a fast initial drop that flattens, and that curve gets misread as a leak.
6. Record the initial pressure P1. Read the pressure at the end of stabilisation as the baseline.
7. Start the measurement window. Isolate the cavity and wait for the configured interval.
8. Record the final pressure P2. Compute the drop.
9. Apply temperature and volume correction. Convert the measured drop to standard conditions, or use the differential method so that temperature effects cancel.
10. Decide and document. Compare the corrected drop with the threshold, issue a verdict, and store the pressure curve.

The centre of gravity of the procedure in one line: filling is preparation, stabilisation is the real work, measurement is just reading. Almost every case of "the airtight test results make no sense" traces back to a stabilisation time that was quietly shortened.

Setting the Parameters: Pressure, Stabilisation, Measurement Time and Threshold

These four parameters constrain each other and must be set as a set rather than one at a time.

Test pressure. The principle is "close to the real service condition without damaging the product". For a case whose purpose is water resistance, the IPX7 static head offers a reference point: one metre of water corresponds to about 9.8 kPa. A test pressure in the range of a few to a few tens of kilopascals therefore has an engineering basis. Note that gas and liquid load the walls differently. Gas is compressible and a thin-walled case bulges more readily under air, so once a pressure is chosen, run a deformation check at that pressure.

Stabilisation time. Empirically it scales with cavity volume. Small cases may need only tens of seconds; large ones may need several minutes. A practical way to confirm adequate stabilisation is to read the pressure twice in succession: if the difference between the two readings is less than a tenth of the allowed drop over the measurement window, the cavity can be considered balanced.

Measurement time. A longer window detects smaller leaks but accumulates more drift. A typical compromise is ten to sixty seconds. Where very small leaks must be detected, the better combination is a longer window together with the differential method rather than simply stretching an absolute single-chamber measurement.

Acceptance threshold. The threshold should not be guessed. Derive it from the distributions of known-good and known-defective parts. Measure the decay of a set of units that already passed immersion testing, then measure a set of units with deliberately induced defects such as a missing gasket segment or an unclamped latch, and place the threshold at the boundary between the two distributions with a safety margin. Only a threshold derived this way carries process meaning.

Temperature and Volume Correction: Where Most Errors Come From

Temperature is the main enemy of airtight testing, for a simple reason: gas pressure is extremely sensitive to temperature. Near ambient conditions, a one degree change in a closed cavity shifts pressure by roughly 0.34 percent in absolute terms. If the acceptance threshold sits at 0.5 percent, then a few degrees of ambient swing consumes the entire signal budget.

Drift has three sources:

• Heat from filling. Gas warms as it is compressed, so immediately after filling the cavity is warmer than its surroundings. Enter measurement too soon and the cooling gas produces a pressure fall that looks exactly like a leak.
• Ambient fluctuation. Air conditioning cycling, doors opening and closing, and changes in sunlight all produce slow drift.
• Part temperature. A case straight from an injection machine, an oven or a wash line has walls at a different temperature from the room and needs time to recover.

There are three corresponding countermeasures:

1. Extend stabilisation so that fill heat dissipates. This is the most direct and most effective fix.
2. Use the differential method so that ambient drift acts on both the part and the reference and cancels.
3. Introduce temperature compensation using a sensor inside the cavity, converting the reading to a reference temperature through the ideal gas relation.

Volume correction matters just as much. What you measure is the total cavity volume, not the marketed capacity of the case. Total volume includes the interior, the fixture cavity, and the dead volume of tubing and valves. Longer or wider tubing raises the dead-volume share. Consequently, any change to the fixture, the manifold length or the tube diameter requires updating the volume parameter in the software; otherwise the threshold silently loses accuracy. This is a particularly common oversight when a line is duplicated at a second plant.

A useful rule of thumb: a leak shows up as a change in the slope of the decay curve, while thermal drift shows up as a slow monotonic change. Reading the shape of the whole curve usually separates the two better than comparing the first and last number alone.

Comparing the Four Families of Leak Test Methods

Pressure decay is not the only route. Knowing the boundaries of each method prevents using the wrong tool.

• Pressure and vacuum decay. Quantitative, fast and automatable, well suited to production screening. Sensitive to temperature, and the detection limit is bounded by the measurement window.
• Bubble emission, ASTM D3078. Immerse the specimen or coat it with soap solution and watch for continuous bubbles. The advantage is that the leak location becomes visible. The drawbacks are subjectivity, no quantification, and residual liquid.
• Immersion, IEC 60529 IPX7 and IPX8. Closest to the real service condition and the reference method for type testing. Drawbacks are the risk of wetting the product, long cycle time, and a partly subjective verdict.
• Tracer gas, typically helium mass spectrometry. Extremely sensitive, capable of finding very small leaks, common in medical and semiconductor applications. The drawbacks are equipment cost, helium consumption and complex tooling.

The most economical engineering combination is pressure decay for production screening, bubble emission for localisation, and immersion for type validation. They are not substitutes for one another; they are stages on a line.

Locating the Leak: Gaskets, Latches, Valves and Inserts

Leaks in protective cases cluster heavily. Five categories account for the overwhelming majority of failures.

First, the gasket interface. Mis-seated gaskets, short shots or flash, incorrect groove fill ratio, and insufficient compression. The gasket is the first line of defence, and its material selection and ageing behaviour are covered in how to choose case seal materials and seal ring design for outdoor cases.

Second, rim flatness. A moulded part that cools unevenly can warp along the rim. Warping over-compresses the gasket in one place and under-compresses it in another. Over-compression causes permanent set; under-compression creates an open path. Check by fitting the lid and inspecting the compression trace around the rim, or by checking contact with a feeler gauge.

Third, latches and hinges. A latch does not exist merely to lock; it exists to compress evenly. A single-sided latch, uneven clamping force or a latch placed off-centre all produce non-uniform pressure around the rim. Selection guidance is in how to choose case latches.

Fourth, the pressure equalisation valve. This is the special case in airtight testing, because by design it lets gas through. The next section deals with it separately.

Fifth, inserts and cable entries. Metal inserts and polymer substrate have different coefficients of thermal expansion, so a temperature cycle can open micro-gaps at the interface. Junction boxes with under-potted cable entries are another classic leak site.

How a Pressure Equalisation Valve Changes the Test

A pressure equalisation valve, also called a breather valve or vent valve, makes airtight testing one layer more complicated than testing a plain sealed box. Its purpose is to allow air to pass slowly when the case experiences temperature change, altitude change or transport pressure differential, while blocking liquid water. Which means that a valve is deliberately leaky in the airtightness sense.

Testing a case with a valve installed directly by pressure decay therefore measures the permeability of the valve, not the sealing performance of the case. This is one of the most frequent sources of misjudged results on a production line. There are three correct ways to handle it:

1. Test the shell on its own. Run the airtight test before the valve is installed, or blank the valve port with dedicated tooling. The result is the sealing capability of the case itself, which is what process control actually cares about.
2. Test the valve separately. Mount the valve on a dedicated fixture and measure its cracking pressure, closed-state sealing and permeability as an incoming inspection item.
3. Test the assembly with the valve isolated. For a finished case, build the port isolation into the sealing head so that the test covers only the shell and gasket.

Valve degradation is the principal long-term risk. Diaphragm ageing, dust or salt crystals on the seat, and UV exposure all turn a controlled breather into a continuous seep. The structural considerations are covered in the role and selection of case pressure valves.

A practical rule: the airtight threshold for a case with a valve is typically about an order of magnitude looser than for the same case without one, and the technical agreement must state whether the test is performed with the valve installed or removed. Otherwise the same case can receive two completely different verdicts depending on the test state.

Putting It on the Line: Full Inspection, Sampling and Traceability

Moving a laboratory method onto a production line comes down to three problems: cycle time, tooling and data.

Cycle time. The floor for one cycle is the sum of fill, stabilisation and measurement. If that exceeds what the line allows for a large case, there are three routes: run multiple test stations in parallel off one air supply, switch to the differential method to shorten stabilisation, or test large cases by sampling while keeping full inspection for small ones. Whatever trade-off you make should be justified in the process document rather than emerging from an unspoken decision to test less.

Tooling. The difference between laboratory and production tooling is repeatability of operation. A lab can align a fixture carefully; a line needs the part to seat correctly the moment it is placed. Design for mistake-proofing: locating pins, hard stops and quick-change sealing heads that squeeze human alignment error toward zero. The fixture's own leak rate must be checked regularly against a blanking plate.

Data. For B2B supply, especially into automotive, medical and defence channels, traceability of airtight test data is routinely audited. At minimum record the model and batch, the equipment identifier, test pressure, stabilisation and measurement times, measured drop, corrected drop, verdict, operator and timestamp. Retention should cover the product warranty period or the customer's stated requirement.

Periodic verification. Three cadences are worth adopting. Every shift, run a two-point check with a blanking plate, which cannot leak, and a calibrated leak orifice of known flow. Weekly, verify against a golden reference, meaning a unit that has already passed immersion testing. After any change to tooling or manifold, recalibrate the volume parameter. This regime separates equipment problems from product problems cleanly and stops blame from landing in the wrong place.

Common False Verdicts and a Troubleshooting Table

When readings look wrong, resist the urge to blame the product first. Working through the table below usually finds the cause faster.

The golden rule of troubleshooting: disprove the equipment before suspecting the product. Run one test with a blanking plate. If the plate "leaks", the problem is in the equipment or the tooling.

Frequently Asked Questions

Q: Can pressure decay replace the IP67 immersion test? A: Not replace, only complement. Pressure decay measures how readily gas crosses the sealing interface and judges by the pressure drop. The IP67 immersion test measures whether liquid water enters the enclosure after thirty minutes at one metre and judges by whether water is found inside. The physical processes differ: gas molecules are far smaller than the clusters that make up liquid water, so a case with excellent airtightness is normally also water resistant, but a pass on airtightness does not by itself imply a pass on IPX7. Water ingress is also influenced by surface tension, water movement and case deformation. The right structure is to position pressure decay as process control and production screening, and to position IEC 60529 immersion as type validation and certification evidence. The best practice that binds them is this: take a unit that has passed immersion as the golden reference, use it to calibrate the decay threshold, screen production against that threshold, and re-verify the equipment against the reference on a fixed schedule.

Q: What test pressure should be used? A: There is no universal number. It is set jointly by the target service condition and the stiffness of the case. From the service side, the IPX7 static head offers a reference: one metre of water is about 9.8 kPa, so a gauge pressure of a few to a few tens of kilopascals has an engineering basis. From the stiffness side, remember that gas loads a plastic case differently from liquid. Gas is compressible and a thin-walled case bulges under air; bulging stretches the rim and changes gasket compression, which can manufacture a phantom leak. The practical way to choose is to raise the pressure in steps while watching both the stability of the curve and the appearance of the case, and to settle on the point that produces neither visible distortion nor an unstable reading. Where the customer technical agreement specifies a pressure, follow the agreement and state the pressure, cavity volume and measurement time in the report.

Q: Why does the same case give a different result on a second test? A: The most common reason is gasket rebound lag. Rubber takes time to recover its section after compression, so if the second test follows immediately, the gasket is still compressed and the decay differs from the initial state. Fix it by specifying a settling interval between tests, or by leaving the case open and relaxed between runs. A second reason is temperature: residual heat from the first fill leaves the cavity slightly warmer than ambient, and a warmer cavity produces a smaller drop. A third is contamination: a trace of moisture or dust on the sealing face can move between runs. To identify which applies, examine the curve shape. A rebound problem gives curves of similar form but different amplitude, while a temperature problem shows a slow monotonic drift.

Q: How should a case with a pressure valve be tested for airtightness? A: Separate the sealing of the shell from the permeability of the valve. A pressure equalisation valve is by design a controlled breather, so running pressure decay on the assembled case measures the valve, not the shell, and hides shell defects in the process. The recommended three-part approach is as follows. First, complete the airtight test before the valve is fitted, or blank the valve port with dedicated tooling, which yields the pure sealing capability of the shell and is the core process control figure. Second, mount the valve on a dedicated fixture and measure cracking pressure, closed-state sealing and room-temperature permeability as an incoming inspection item. Third, for the assembled case, build port isolation into the test head. Note also that the acceptance threshold for a valved case is typically about an order of magnitude looser than for the unvalved version, depending on valve specification, and the technical agreement must state whether the test is run with the valve in place or removed.

Q: How should the acceptance threshold be derived scientifically? A: The sound method is to place the threshold at the boundary between two distributions. First, take no fewer than twenty units that have already passed IEC 60529 immersion testing and measure their decay under the specified conditions to establish the distribution for good parts. Second, create units with known defects, for example a gasket with a segment omitted, a latch left unclamped or a deliberate score on the rim, and measure their distribution in the same way. Third, set a preliminary threshold at the boundary between the two distributions and then move it toward the good-part side by a margin of two to three standard deviations. A threshold obtained this way is neither loose enough to pass defective units nor tight enough to reject good ones. Remember that the threshold is bound tightly to cavity volume; if the volume changes, the threshold must be recalibrated, and the same applies after tooling changes, longer manifolds or sensor replacement. Writing the threshold, the volume and the test conditions into the process document together is what produces consistency across shifts and across plants.

Q: How small a leak can pressure decay detect? A: It depends on measurement time, sensor resolution and cavity volume together, and there is no single figure. In principle the pressure drop equals leak rate times time divided by corrected volume, so a longer window or a better sensor both improve detection. In practice two factors dominate. The first is thermal drift: over a long window, accumulated drift gradually buries the real signal. The second is cavity volume: the larger the volume, the smaller the drop produced by the same leak. A small case tested with the differential method can detect very small leaks indeed. For a large case, simply extending the measurement time runs into diminishing returns, and periodic verification by immersion or tracer gas becomes the more sensible route. The engineering recommendation is to start from the defect you actually need to intercept, use deliberately defective parts to establish the required detection level, and work backwards to the method and parameters rather than choosing parameters first and asking afterwards how small a leak they can catch.

Q: How often does airtight test equipment need calibration? A: A three-tier regime works well: daily, weekly, and after any change. Every shift, run a two-point check using a blanking plate to confirm the zero-leak baseline and a calibrated leak orifice to confirm the response to a known leak, and require both readings to fall within tolerance. Weekly, verify with a golden reference, meaning a case that has already passed immersion testing, to confirm that its decay reading still falls inside the expected band; this prevents the threshold from silently expiring as equipment drifts. After any change to tooling, tubing, sensors or software, recalibrate the cavity volume and revalidate the threshold. In addition, send pressure and temperature sensors for periodic metrological calibration, and watch the zero drift of differential sensors specifically. Keeping a record of every check is not only a quality management requirement but also the evidence you will need during customer audits and disputes.

Q: Why does a case that passes airtight testing still let water in? A: Three explanations cover most cases. First, the threshold is too loose: it was set by judgement rather than validated against a defective-part distribution, so a real but small leak slips through. Such a leak may barely register as a pressure drop but can still admit water over thirty minutes. Second, the defect path is unusual: some defects only appear under liquid conditions, for instance where the gasket contacts the rim well under air but separates microscopically under water pressure because of surface tension effects and case deformation. Third, the test states differ: the airtight test was performed with the valve removed, the case open for inspection, or the foam absent, while the immersion test used the fully assembled product. The suggested order of investigation is to review where the threshold came from, use bubbles to localise the suspect unit, and then confirm that the two tests used strictly identical sample states, with those states written into the technical agreement.

Q: Can a small buyer test airtightness without proper equipment? A: Yes, provided the accuracy limits are understood. For small volumes or on-site acceptance, the common improvised methods are immersion to look for continuous bubbles, soap solution applied to sealing faces while the case is pressurised, and placing a desiccant pack with a humidity indicator card inside the case and observing it over time. These cost almost nothing and will reveal gross defects, but they cannot produce a quantitative criterion, they struggle with small leaks, and they are subjective. If you use them, do at least three things: standardise the operating conditions including water temperature, soak time and whether pressure is applied; keep a known-good case as a comparison reference; and state clearly that the method is only an on-site pre-screen and cannot serve as a certificate of conformity or a basis for dispute. The detailed procedure and its limitations are covered in how to test case airtightness yourself.

Conclusion and Related Reading

Back to the question in the title: how do you perform an airtight test on a protective case? The answer is to build a process control chain with pressure decay at its centre and immersion testing as its benchmark. Fill, stabilise and measure form one cycle. The pressure drop is the criterion. The differential method and temperature compensation remove environmental interference. A three-level verification regime of blanking plate, calibrated leak orifice and golden reference keeps the equipment honest. There is nothing exotic about any of it. It is a parameter system that has to be executed with discipline.

Three actions you can take immediately. First, fix the five parameters of test pressure, stabilisation time, measurement time, cavity volume and acceptance threshold, and write them into the process document, because consistency across shifts is worth more than peak accuracy in any single run. Second, derive the threshold from the two distributions of immersion-passed units and deliberately defective units, so the number carries process meaning rather than being a guess. Third, state the valved and unvalved test states explicitly in the technical agreement, and fold the daily blanking plate check and the weekly golden reference verification into routine shift work.

JUNZHJIA, manufactured by KeXin New Materials (Guangdong) Co., Ltd., produces protective cases, tool boxes, military-spec storage cases and waterproof junction boxes for wholesale, distribution, OEM and ODM programmes and global supply. The company can configure airtight test schemes, sealing fixtures and data templates to match a customer's service conditions and acceptance standards, and supplies structural documentation, material data and test files alongside the product.

Related Reading

• How to Test Case Airtightness Yourself: Simple Methods and Their Limits
• How an IP67 Protective Case Is Engineered: Structure and Sealing
• How the IP67 Submersion Test Is Run: Depth, Duration and Verdict