12.1 Leak Testing & Acoustic Emission

Leak Testing (LT)

Leak testing (LT) detects and, in some techniques, locates or measures through-thickness leakage across a pressure-retaining or vacuum boundary. A leak is a physical passage that lets fluid flow from high to low pressure, so LT is fundamentally a test of system tightness, not of a solid-material discontinuity. Unlike VT, PT, MT, ET, UT, and RT, which reveal a flaw in the wall itself, LT tells you whether the boundary holds. Because of this, a Level III chooses LT whenever the requirement is expressed as a leak rate, a pressure hold, or fitness of a sealed system.

Leak rate and units

The measured quantity is leak rate, the product of pressure and volumetric flow per unit time. Common units are std cm3/s (also written atm cc/s), mbar-L/s, and the SI Pa-m3/s. Useful equivalences: 1 atm cc/s is about 1 mbar-L/s and roughly 0.1 Pa-m3/s. Sensitivity requirements drive technique choice, spanning coarse (about 1E-3 std cc/s) to ultra-fine (below 1E-10 std cc/s).

Three families of LT technique

1. Pressure-change testing. Governed by the gas law under isothermal conditions, this family includes pressure decay (pressurize, isolate, watch pressure fall), pressure rise, and vacuum decay (evacuate, watch pressure rise). It measures the whole system but cannot locate the leak. Temperature compensation is critical: because pressure tracks absolute temperature, a fraction of a degree of drift can masquerade as a leak, so a stabilization period precedes the hold and a sealed reference volume is often compared against the part. Typical field sensitivity is on the order of 1E-3 to 1E-4 std cc/s.

2. Bubble testing. The part is pressurized and a film solution is applied, or it is immersed, and escaping gas forms visible bubbles at the leak. Variants include direct-pressure immersion, solution-film brushing, and the vacuum-box technique for one-side access such as tank bottoms and weld seams. Bubble testing locates the leak but only resolves relatively gross leaks (about 1E-3 to 1E-4 std cc/s); a surfactant lowers surface tension so smaller leaks bubble. Guidance appears in ASME Section V, Article 10 and ASTM standards such as E1066 (immersion) and E515 (film).

3. Tracer-gas testing. A tracer gas is introduced and a detector responds to it. The halogen diode detector senses halogenated tracers, while the helium mass spectrometer leak detector (MSLD) is the workhorse of high-sensitivity work, reaching roughly 1E-9 to 1E-12 std cc/s. Helium is preferred because it is inert, non-toxic, very small (permeates tiny channels), and present at only about 5 ppm in air, giving a low background that a mass spectrometer can pick out. Modes include hood (total), sniffer (locating), and vacuum/spray.

Acoustic Emission (AE)

Acoustic emission (AE) is fundamentally different from every other method in this guide: it is passive. Rather than injecting energy (as UT injects sound or RT injects radiation), AE listens for transient elastic stress waves that the material itself releases when a discontinuity grows or deforms. Sources include crack propagation, plastic deformation, fiber breakage in composites, corrosion, phase changes, and leakage. Because emissions occur only when a flaw is active, the structure must be stressed during the test, typically by a proof load or pressurization cycle.

Sensors and source location

AE uses piezoelectric sensors coupled to the surface, spread across the structure. When a source emits a burst, each sensor records a slightly different time of arrival; comparing arrival times across an array lets the system triangulate the source location (linear location for pipe, planar for vessels). This is why AE is called a global or volumetric-monitoring method: a handful of sensors watches an entire pressure vessel at once, instead of scanning point by point like UT or MT.

What AE tells you and its limits

AE excels at finding active, growing flaws during a load test and ignoring benign ones, so it is used for proof-testing vessels, monitoring in service, and detecting incipient failure. The Kaiser effect (no new emission until a prior maximum load is exceeded) and the Felicity ratio (early re-emission indicating damage) are key interpretive concepts. As a worked idea: if a vessel emits activity well below its previous peak load, the Felicity ratio is less than one, signaling that damage has advanced since the last test. Limits: AE locates and monitors but does not size or characterize a flaw; it needs stressing; it is sensitive to background noise; and a located source must be confirmed with another method such as UT or RT.

Choosing between LT and AE

Although both are grouped as "other" methods, LT and AE answer opposite questions. LT asks whether the boundary holds, quantifying or locating a through-leak in a sealed or pressurized system, and is the right tool when the acceptance criterion is a leak rate. AE asks whether the structure is degrading, flagging active cracking or plastic strain during a controlled load and pointing to where to inspect next. A common mistake is treating AE as a standalone accept/reject tool; because it neither sizes nor images a flaw, its located sources are triaged and then confirmed with UT, RT, or PT. Likewise, a whole-system pressure-decay result that fails must be followed by bubble or tracer testing to pinpoint the leak.