What is the minimum leak rate that can be detected reliably using airborne ultrasound?

There is no universally applicable minimum leak rate for airborne ultrasonic testing.

Ultrasonic testing does not directly detect the size of an opening. Instead, it detects the turbulent flow generated when gas escapes from or enters a system through a leak. Whether this produces a sufficiently strong ultrasonic signal depends on several factors:

• the pressure difference between the system and its surroundings,
• the type and temperature of the gas,
• the size, shape, and surface condition of the leak,
• the flow conditions at the leak,
• the measurement distance and orientation of the sensor,
• the directivity and sensitivity of the testing device,
• ultrasonic background noise in the surrounding area.

Very small leaks or low pressure differences may not generate sufficient turbulent flow. The resulting signal may also be below the background level and therefore cannot be detected reliably.

In leak testing, the leak rate is often expressed in mbar·L/s (millibar litres per second). This unit describes the amount of gas flowing through a leak per unit of time as a pressure-volume product.

In practice, a leak rate of approximately 10⁻² mbar·L/s is sometimes used as a guideline for the transition range in which localization using airborne ultrasound may be possible under suitable conditions. However, this value is not a guaranteed or generally reproducible detection limit.

A tightness requirement of 10⁻² mbar·L/s, sometimes described informally as “watertight,” cannot be verified reliably and reproducibly using conventional airborne ultrasonic equipment under normal industrial conditions.

Airborne ultrasonic testing is very well suited for locating economically relevant leaks in compressed air, gas, and vacuum systems. However, if compliance with a specified leak-rate limit must be demonstrated, the testing method must be validated under the actual test conditions. Other leak-testing methods may be required for very low specified leak rates.

Locating a leak must also be distinguished from quantitatively determining the leakage loss.