A component is regarded as being impervious in the physical sense if a previously specified test method of a certain sensitivity commensurate with the purpose is unable to detect the passing of a suitable test medium from one chamber into another chamber or into the atmosphere. Frequently used methods are the liquid leak test, the bubble leak test, and the gas leak test.

Liquid leak test.

The idea underlying the liquid leak test is that a test specimen holding pressurized liquid will show a leak by the emergence of the liquid. It is customary to conduct this test simultaneously with the water pressure test stipulated in the code of regulations for pressure vessels. All openings in the test specimen must be rendered liquid-tight beforehand. When filling the specimen, it must be carefully vented of all air, as any leaks in the area of air or gas cushions will be unable to make themselves felt. A pressure gauge is connected to the test chamber to measure the testing pressure, which varies according to the specimen's design data. At all events it should be at least as high as the ultimate operating pressure. The precautions stipulated in the pressure vessel regulation must be observed for the test. After applying the testing pressure and waiting for an appropriate length of time, the specimen must be examined carefully for damp areas. A guide value for the detection sensitivity achieved by this method is 0.5 mbar-1/s. This method used as a static leak test on large mechanical seals is usually unable to detect a leak, as it is generally impossible to view the inside of the mechanical seal, e.g. its sliding faces, in the assembled state. A leak is indicated by the pressure drop per unit of time, which is thus a means of assessment.

Bubble leak test. In the immersion bubble test, the test specimen is filled with air or a different gas and immersed in a water bath under pressure. A leak is shown by rising bubbles of gas. The detection sensitivity of this method lies in the magnitude of 10^-4 mbar/1/s. In the bubble method using foaming liquid, the test specimen is again pressurized by air or a different gas, but a foaming liquid (e.g. Nekal^® solution) spread over the test area is used as indicator. If there is a leak, more or less fine clusters of foam develop over it. The detection sensitivity of this method also lies at around 10^-4 mbar·l/s.

Gas leak test. The three most frequently used test methods are

• the helium leak detection method,
• the halogen leak detection method,
• the SF₆ leak detection method.

All three test methods are similar in their procedures, differing solely in the type of test gas and the mode of operation of the detection cell. The idea underlying the gas leak test is to use a suitable detector to measure the quantity of gas escaping through a leak. In the case of the helium leak detection method, a mass spectrometer is used to measure a current of ions that is directly proportional to the number of gas molecules. Under ideal conditions it is then possible to specify exact leakage rates. A distinction is drawn between the vacuum mode, which under ideal conditions can attain a detection sensitivity of 10^-11 mbar·l/s, and the pressure method, which is capable of achieving a detection sensitivity of just 10^-6 mbar·l/s.