TechArchive

Original article date: October 1998

When should a leak test be performed? On line? Post production?

Sophisticated electronics and computerised systems have revolutionised the testing of many components yet leak testing often remains crude and perhaps not well understood. But one thing is certain: the cost to industry of leaks taking into account detection measurement prevention and correction is massive.

Leak detection says Brian Wilkie of TQC is very much a horses for courses scene. There are no convenient or general answers to simple questions. The commonly used test techniques of leak measurement are: differential pressure continuous flow and gas trace. There are multiple divisions within each involving either gas positive pressure vacuum or combinations.

Wilkie says that the common feature of leak location methods is that they are almost always operator-dependent require strict supervision and are very often messy! No easy task then. Common methods include:

Observation of bubble streams

Chemical trace

Chemical penetration

Gas sniffing

Ultrasonic sensing

The first decision to make is “to what level of leak tightness” – how small a hole is it necessary to find? But a hole needs to be defined in terms of how it will affect performance. For example a 1um hole in a rubber gardening glove is insignificant – the same hole in a surgeons glove could allow a serious infection. A valve that allow 5ppm of non toxic gas to escape may not give cause for concern but soon would if the gas was radioactive Therefore the first step in any leak detection specification is to define the performance characteristics of the component to be tested and a fitness for purpose leak specification.

The criteria for fitness for purpose can change however. For example over the last two years environmental kegislation has been introduced to reduce pollution from automotive fuel components. In basic terms the fuel system must now be leak tight to fuel vapour rather than just liquid fuel. Naturally this means that leak testing has to be more stringent and leak rate specifications are being lowered from to 1 x 10-3 cc/min or lower. According to Janet Grimwade at Ion Science these new standards cannot be met by traditional techniques such as pressure decay. The automotive industry is therefore moving to more sensitive tracer gas techniques such as helium mass spectrometry or the Photec system from Ion Science. Whilst highly sensitive mass spectrometry can be relatively expensive (it is about three times the cost of pressure decay) and can be complex to operate and maintain. Photec claims to be more sensitive than pressure decay but with lower capital and maintenance costs than helium mass spectrometry

Metallurgy television semiconductors electronic coatings all use vacuum within their manufacturing processes. Ensuring that there are no leaks in the vacuum is a critical factor in quality control. In the past the only way to leak check a vacuum system was to use a helium mass spectrometer. Whilst eminently accurate and sensitive mass Spectrometers are complex to use and not particularly sportable. Ion Science is therefore about to launch the Vac-Check. This light hand-held the instrument will provide a convenient alternative to mass spectrometers for routine vacuum process monitoring.

The effects of temperature can affect the accuracy of automatic leak detection equipment. Small 0.1Cdeg rises in the temperature of a component cause the air inside to expand. If this occurs during test it can appear that the component is actually filling not leaking leading to a false “pass”. Conversely you can get a false “fail” from a component cooling down.

Products which are designed for heat transfer applications such as heat exchangers radiators and condenser and evaporator coils are the most susceptible to the effects of temperature.

Considering that a 25um hole can cause a car radiator to fail the detection of the smallest leaks those most likely to produce false “pass” results is crucial. AI Qualitek’s latest versions of its Q120 series the Q620T models include patented electronics which compensate for the effects of temperature and correctly identify leaks even when the component is the subject of random temperature variations.

According to Janet Grimwade of Ion Science it is inportant to leak test as early as possible in a production cycle for the obvious reason that there is not much point in completing the assembly process of a leaking component. However this is not always possible or logical as an assembly process can cause a leak in an otherwise sound component. It is therefore generally advisable to have a leak test at the beginning and end of a production cycle where it is economic to do so.