Home / Blog / Waterproof Technology

The Air Inflation Leak Test

What an air inflation test can actually detect

The test is designed to reveal through-paths in an otherwise enclosed bag. Air may escape through a film pinhole, a narrow channel across a weld, a poorly seated valve, a cracked fold line, or a closure clamped incorrectly by the fixture. Large openings produce an audible hiss and fast pressure loss. Very small defects may show only as a growing foam crown under bubble solution. Location matters: a bubble line along one tool edge points toward weld setup, while isolated bubbles in the panel suggest material damage or contamination.

A pass means no leak was observed under the stated pressure, duration, preparation, and detection method. It does not mean permanent airtightness or unlimited waterproofing. Flexible bags breathe slightly through some polymers and fixtures can lose air without the product leaking. Temperature also changes pressure. Treat the result as one controlled screen within a wider programme that includes visual inspection and, where required, immersion testing. Clear scope prevents a 45-second bench check from becoming an unsupported lifetime claim.

Pressure difference drives every leak signal

Air flows because pressure inside the bag is higher than pressure outside. Increasing that difference makes a small path easier to detect, but it also loads every square centimetre of fabric and can distort the product. At 2 kPa, each 100 cm2 panel area experiences about 20 newtons of separating force. Across a broad bag wall, the total load becomes substantial even though the gauge number looks small. That is why low-pressure fixtures require a relief device and why hand pumps are preferable to unrestricted compressed-air lines.

The relationship is not perfectly linear because flexible seams open, folds move, and foam changes the path. A channel that remains closed at 0.5 kPa may open at 1.5 kPa, then remain visible after pressure drops. Conversely, an overinflated bag can create a temporary crease leak that never occurs in service. The pressure differential guide explains load distribution in more detail. A useful method therefore sets a pressure window, not merely a vague instruction to inflate firmly.

Equipment for a controlled low-pressure check

A basic station needs a low-volume hand pump or regulated source, an adaptor suited to the valve or temporary inlet, a gauge readable to at least 0.1 kPa, a shut-off valve, a pressure-relief valve, a timer, and a non-marking support surface. Add good task lighting, a washable tray, absorbent cloths, and a skin-safe bubble solution. The operator also needs a blunt removable marker to circle defects. Every hose connection must be secured so movement does not create an apparent product leak.

Choose a gauge range close to the working pressure. A 0 to 10 kPa gauge gives useful resolution around 2 kPa; a 0 to 700 kPa workshop gauge does not. Verify zero before each shift and compare the station with a reference gauge on a planned interval. The relief valve should open below the pressure known to stretch seams or fixtures. Never hold a bag against the face while pressurising it, and keep the volume low enough that an accidental release has little stored energy.

Preparing the bag and the test station

Condition the bag at a stable room temperature for at least two hours when results will be compared across lots. Empty every pocket, remove loose inserts, and inspect the surfaces dry. Record cuts, scratches, creases, contamination, and previous repairs before pressure is applied. Close zips and valves according to their normal instructions. For a roll-top, decide whether the closure itself is under evaluation or whether a temporary sealed inlet isolates the welded body. Mixing those two aims produces confusing results.

Run a blank check on the station by sealing the adaptor without a bag, applying the target pressure, and watching for 60 seconds. Any meaningful drop belongs to the fixture and must be corrected first. Place the test item on a smooth cradle so its seams remain visible and unsupported corners do not carry the full load. Keep direct sun, fans, and heaters away; a temperature change of only a few degrees can shift the gauge enough to resemble a slow leak.

The step-by-step air inflation leak test

Connect the adaptor, confirm the relief path is open, and introduce air slowly. Stop at the lower edge of the approved pressure window, allow 5 to 10 seconds for the film to settle, then adjust once to the target. Close the inlet and start the timer. Sweep the base, longitudinal seams, corner transitions, attachment patches, valve, and closure in the same order every time. Listen first in a quiet station, then apply bubble solution only to locations that need higher sensitivity. Avoid pressing on the body during the timed hold.

At the end of 30 to 60 seconds, record final pressure and every visible signal before marking the bag. Release air through the valve or adaptor rather than opening a loaded roll-top suddenly. Wipe away solution with fresh water and dry the surface before diagnosis. If a defect appears, repeat at the same setting once after checking the fixture. A second result that emerges at the same point is far more useful than repeatedly changing pressure until some bubble can be produced.

1. Verify gauge zero, relief operation, timer, and the blank fixture.
2. Condition, identify, and visually inspect the dry bag.
3. Connect the lowest-volume suitable adaptor and secure every joint.
4. Raise pressure gradually to the approved 1.0-2.5 kPa window.
5. Hold for 30-60 seconds without squeezing or moving the bag.
6. Inspect in a fixed route, using bubbles only where needed.
7. Record, depressurise safely, rinse, dry, and segregate any failure.

Listening, touching, and gauge decay

Audible inspection is quick but depends on background noise and operator hearing. A gross channel may hiss from several centimetres away; a fine pinhole may remain silent. Move the ear near the surface only after inflation has stopped and keep the face out of the release path. A light hand held above the suspected point can feel a larger jet, but touching the fabric may close a tiny route or change panel pressure. Use sound and airflow as location aids, not as the sole acceptance method.

Gauge decay looks objective yet includes the entire system: product, hose, adaptor, temperature, and gauge resolution. Define a stabilisation period and compare against a verified blank. For example, a method may allow no more than 0.2 kPa loss from a 2.0 kPa start over 60 seconds after 10 seconds of settling. That number must come from validation with known good and deliberately defective samples. Copying it into an unrelated bag size or material can reject sound products or miss small leaks.

Using bubble solution without creating false results

A bubble solution lowers surface tension so escaping air forms visible domes instead of disappearing. Apply a thin continuous film with a soft brush or squeeze bottle, beginning at the lowest point so runoff does not cover unchecked areas. A true leak creates bubbles repeatedly at one origin after the surface foam is cleared. Pre-existing froth that does not regrow is application foam, not evidence. Photograph the signal beside a scale and mark the dry reverse side once pressure has been released.

Use a mild formulation that has been checked against the fabric, print, adhesive, and hardware. Concentrated detergent can leave residue, alter gloss, attack an unprotected coating, or creep beneath labels. Rinse tested zones with fresh water and dry them open. For regulated laboratory work, document solution identity, concentration, application amount, and dwell time. Test-method development principles from ASTM reinforce why repeatable apparatus and a defined observation rule matter more than producing dramatic foam.

Reading seam and corner failure patterns

Straight-seam leakage often points to a local contaminant, tool pressure variation, or insufficient energy. A channel at the end of a weld may indicate poor run-off geometry. Repeated bubbles around a tight inside curve suggest wrinkling or uneven overlap rather than weak base material. Base corners are especially informative because several force directions meet there. Record the distance from a fixed datum so technicians can compare the defect with tooling position, material joins, and the sequence used during assembly.

Do not seal the visible point immediately and return the bag to use without identifying the pattern. Cut a failed sample only after its location and pressure result are documented. A cross-section can reveal missing polymer contact, foreign matter, excessive squeeze-out, or a narrow bond. Compare with retained good seams from the same material lot. The seam failure guide explains how small channels propagate under flex, while the air test supplies the location needed for useful root-cause work.

Testing valves, closures, and attachment patches

Valves need separate attention because leakage can occur through the core, beneath a gasket, or around the welded flange. Close the valve normally, wet each interface in sequence, and observe before touching it. If bubbles appear beneath the cap, inspect the seat and seal. If they begin at the flange edge, review weld continuity. Torque should follow the approved design; tightening until bubbles stop can damage threads and hide an assembly issue rather than correcting it.

Roll-top closures are difficult to quantify with inflation because the fold is both seal and test boundary. Use a consistent fill level, three folds, buckle orientation, and clamping arrangement. Attachment patches deserve a complete perimeter sweep because webbing can hide a channel. If the purpose is body integrity, isolate the closure with a dedicated fixture. If the purpose is real-use closure performance, combine low-pressure air screening with controlled roll-count testing and a suitable water exposure.

False passes, false failures, and method limits

A false pass occurs when pressure is too low, hold time is too short, solution misses the defect, or a fold closes the leak path. A false failure occurs when the adaptor leaks, temperature falls, solution foams on application, or the operator presses a seam open. Porous accessories can also produce harmless pressure drift when they are outside the waterproof compartment. Method validation should include known defects at the smallest size the process intends to catch, plus sound bags tested by several operators.

The air test cannot establish resistance to sustained water pressure, impact in surf, capillary entry through a dirty closure, or ageing after years of UV and flex. It also does not measure hydrostatic head of a flat material directly. The IEC defines ingress-protection test conditions for enclosures, illustrating why a named exposure and duration are essential. Use air for sensitive leak location, then use the intended water test for the performance claim being made.

Sampling plans at an inspection bench

Testing every bag may be justified when contents are high value or when the station is integrated into assembly. Otherwise, use a documented sampling plan tied to lot size and inspection level. At minimum, include first-off pieces after setup, samples at regular intervals, the final piece before a material change, and any item associated with a process interruption. Segregate retested bags from first-pass bags in the records. A sample plan controls inspection effort; it does not permit an obvious defect elsewhere in the lot.

At an SGS-verified 4,950 m2 production facility with 7 lines and 149 machines, a low-pressure station must use the same gauge range, fixture revision, and work instruction across shifts. Production has run since 2014, with a usual minimum of 500 pieces and sample preparation in 6 to 10 working days. Our production team provides broader development context, while each air-test record should still identify operator, bag code, time, pressure, hold, result, and defect position.

Writing pass criteria and traceable records

A complete instruction states conditioning, fixture, inlet route, pressure window, stabilisation time, hold time, inspection sequence, detection solution, allowable pressure loss, bubble rule, and safe depressurisation. Avoid subjective phrases such as adequately firm or no major leak. A clearer rule might require 2.0 ± 0.2 kPa after stabilisation, a 60-second hold, no repeating bubble growth, and no more than validated gauge loss relative to the blank fixture. The limit must be established for that bag family.

Records should connect each result to material and process history without becoming burdensome. Capture date, station, operator, style, lot, serial or sample number, initial and final pressure, duration, temperature if controlled, detection method, result, defect code, location, and disposition. Use a simple drawing to mark positions consistently. Quality-system concepts published by the ISO support controlled documents and traceability, but the local form should remain usable enough that operators complete it accurately every time.

What to do after a bag fails the air test

Stop, identify, and segregate the item before changing any setting. Confirm the fixture with a blank, then retest once at the original conditions. If the leak repeats, photograph and mark it before releasing pressure. Review nearby bags according to the inspection plan and look for a shared position, time window, material roll, or tool setting. One random panel puncture calls for handling review; five seam channels at the same curve call for a process correction and expanded containment.

Repair only when the product specification permits it and a validated method exists. Then repeat the full air check and any required water exposure rather than testing only the patch. Feed failure data back into weld parameters, material inspection, fixture maintenance, and operator training. The waterproof bag quality-control guide connects this result with incoming and final checks. Used carefully, the air inflation leak test is fast and highly sensitive; used without pressure control or scope, it creates confidence that the evidence cannot support.