Today’s manufactures have a variety of different technologies to choose from when leak testing their products. These include pressure decay systems, mass flow systems, helium or hydrogen accumulation systems, helium or hydrogen sniffers, mass spectrometers and residual gas analyzers.

For all their differences, though, every leak test system has one thing in common: The products being tested need to be effectively sealed and fixtured. Failure to do so can lead to all kinds of problems down the road, especially in an automated production system.

With this in mind, there are a few basic principles that need to be adhered to no matter how sophisticated the actual leak detection method. For example, engineers should test an assembly in the same "direction" as the pressure gradient the assembly will be facing when in operation. This means that if a product will contain a liquid or gas when in use-examples, include everything from automotive fuel tanks and radiators to refrigeration coils and medicine delivery systems-then the testing system will need to create a high-pressure region within the assembly and then monitor the product for any escaping gas.

If, on the other hand, the product needs to remain impermeable with respect to outside gasses or liquids, the test will be conducted the other way around. Examples of products in this category include sealed electronic packages, vacuum vessels or systems, and some types of manifolds or valves.

In many cases, the reason for this approach is obvious: Submitting a product to forces it isn't designed to withstand could easily result in its being ruptured or crushed. However, another important reason for testing in the same direction is that you don't want to "mask" the effects of a component's more subtle, or borderline defects.

Let's say, for example, you are testing the welds on a coil for an air conditioner unit. Pressing "in" on the assembly could also compress the welds, causing them to block the escape of the tracer gas being used, even if they are of questionable quality. Pressing "out," on the other hand, will tend to open up those same welds, quickly exposing a defective part.

Along these same lines, masking is another thing that needs to be kept in mind when thinking about fixturing. It's important to ensure that the part or parts are stabilized. But, that doesn't mean clamping the part down so firmly it can't flex and expose any defective joints. Similarly, you don't want the nesting itself to operate as a kind of plug, stopping up, and therefore masking, any potential leaks.

Finally, it is vital that an adequate seal be created between the assembly and the leak detector. In terms of the different options available, engineers can create a seal around the outside or the inside of a port leading into the assembly. Or, if there is no port, they can create a "face seal" along the same surface where the component will mate with the rest of an assembly. The latter arrangement may require the use of hydraulic clamps, because larger sealing areas require higher clamping pressures. The decision to go with an inner-diameter (ID) or outer diameter (OD) seal is often dictated by part geometry. Nonetheless, all other things being equal, an OD is preferable in that it frees up the entire orifice for either pumping in a tracer gas or pulling a vacuum.

No matter what the seal type, it is important that seal be capable of holding up under the rigors of repeated use. In addition, it's important to ensure that the sealing material is compatible with the tracer gas being used. Running an air-based leak test doesn't pose any undue restrictions. However, materials like silicone can be highly porous to tracer gasses like helium and hydrogen.

Ultimately, as is the case with so much in manufacturing, the key to effective, trouble-free leak testing is to not wait until the last minute. Design-for-assembly is all the rage these days-at least in theory-and these same lessons apply to leak testing. Granted, creating a product that is easy to hook up to a leak detector is probably the last thing on many engineers' minds. But you may be glad you took the time to do so afterwards.