Leak testing is an essential part of product quality assurance in a wide range of industries. Because nothing is ever absolutely leakproof, leak testing basically ensures that the product has been tested for maximum allowable leakage. The process can ensure that flammable, toxic, or corrosive substances are safely confined within the container's walls. It can also ascertain that a liquid or gas essential to a system such as brakes, an air conditioning unit, or hydraulic valves will remain where it belongs for the desired amount of time.
Dunking and Decay
For years, the simpler methods—water dunking and pressure decay measurement—have been the most popular. Each offers the advantage of minimal investment, but each has its drawbacks as well.
Water Dunk. Dunking an object into water can be effective in determining whether and where an object has a leak. Theoretically, bubbles form at the source of the leak and the number of bubbles per minute can signify its size. But a very small leak might make few or minuscule bubbles. If the leak is within a recess, air can collect and remain inside the cavity. Or the bubbles may cling to the test object instead of rising to the surface. Water dunking is also extremely dependent on the operator's involvement. When an object is manually dipped into water, it can pull down bubbles from the ambient air that mask the bubbles caused by a leak. The tester has to wait until the object has cleared itself of non-leak-related bubbles. And a small leak on the far side of the object might not be observed. Finally, water can corrode or otherwise damage the device under test.
Pressure Decay. This technique detects a drop in pressure that indicates a leak. The greater the drop, the larger the leak. This method is both dry and easily automated. But it cannot pinpoint the source of a leak, and its accuracy is affected by a number of variables, among them volume, materials, and temperature. An object with a large volume takes too long a cycle time to be practical; flexible plastic bottles and rubber parts counteract pressure decay by reducing their volume. The temperature inside a test object rises with rising pressure, requiring a wait for stabilization. Handling and ambient conditions can also raise or lower the temperature of the device under test and skew the results.
The use of tracer gases for leak detection has caught on in recent years, with helium leading the pack.
Helium. Helium, the lightest of the inert gases, is present in the air in only very small quantities (~4 ppm). Helium mass spectrometers are extremely sensitive to trace amounts of this gas. These instruments are typically equipped with an external pump that creates a vacuum outside the object of interest and detects the gas as it escapes. But mass spectrometers are delicate, expensive to maintain, and best kept in the lab. Furthermore, helium is viscous and can be hard to clean from the test equipment.
Hydrogen. Hydrogen has suffered from a scary reputation ever since the Hindenburg met a fiery end. It is indeed highly flammable, but can be rendered inert by diluting it with nitrogen. Standard hydrogen-nitrogen mixtures are in common use as shielding gases for welding. According to ISO10156, any hydrogen/nitrogen mixture containing <5.7% hydrogen is classified as nonflammable. For leak-testing purposes, most gas suppliers provide a 5% hydrogen–95% nitrogen mixture.
The lightest known element, hydrogen has half the viscosity of air or helium so it spreads swiftly throughout the test object, penetrates leaks more easily, and vents away more readily than any other tracer gas. And hydrogen detectors cost much less than most mass spectrometers.
The development of a new type of hydrogen detector based on microelectronic sensors offers both high sensitivity and high selectivity to hydrogen. The devices are also robust enough for industrial use, and can detect leak rates down to 5 × 10–7 cc/s. They don't need an engineer to operate them either. The gas is simply injected into the test object, and a hand probe connected to the detector is used to search for leaks (see Figure 1). An audio signal increases in frequency as the probe tip nears the leak location. Hydrogen leak testing can also be automated. In Japan, automotive component manufacturers are beginning to test their products on chamber-type, automated testing systems that save on cycle time and money.
Figure 1. With Sensistor's hydrogen-nitrogen-based leak detector, rugged enough for use on the plant floor, no vacuum is required and nothing gets wet. Here, the gas mixture has been pumped into an automobile engine, and the "sniffer" checks for sources of leakage.
It's clear that there are many options available for leak detection and measurement. But for ease of use and cost effectiveness, hydrogen is hard to beat.
Dr. Claes Nylander is President, Sensistor Technologies, Inc., Linkoping, Sweden.