Manufacturing engineers use leak testing equipment to verify the integrity of products ranging from tiny medical devices to huge aerospace fuel tanks. However, as the need to increase production rates and reduce costs has increased, the degree of accuracy and sensitivity demanded has become higher. In addition, new materials and new products are creating numerous opportunities and challenges for testing critical components.

Fortunately, new technology is being developed to increase throughput, improve reliability and shorten test cycle times, while addressing market demands for lower cost, simplicity, better resolution and more extensive validation. As a result, more and more components are now being leak tested to ensure product quality.

"Parts that were never tested in the past are now 100 percent tested," notes Drew Wilson, P.E., director of technology and business development at ATS Test Systems (Woodbridge, ON), a division of ATS Automation Tooling Systems Inc. (Cambridge, ON). "Most leak testing is now very inexpensive, fast and reliable. Gauge repeatability and reproducibility studies of less than 10 percent are the norm."

However, despite increased demand, Wilson claims there is still a huge lack of basic understanding in the marketplace. "I am amazed at the number of manufacturers that still use bubble testing under water to establish product quality-either in production or to set engineering specifications," he points out. "We are still converting industries from bubble testing to mass flow. There is less and less engineering capability on new projects, and a tendency in many plants to do leak testing the same old way."

Testing Techniques

Leak testing verifies the integrity of a manufactured product. Tests range from pressure testing and pressure decay to destructive burst testing and bubble break point testing. Testing equipment can be used to detect a wide variety of problems and flaws, such as gross leaks, fine leaks, blockage, cracks in seals, crimped lines, rolled seals, kinked hoses, and defective valves and pressure regulators.

There are two basic methods used to test products for leaks. In one method, the product is pressurized, isolated from the pressure source, and observed to detect escaping gas. Bubble, sniff, pressure decay, differential pressure, and mass flow testing fall into this category. In the other method, the product is evacuated and then monitored to see if gas flows into it. Pressure rise and mass spectrometry are in this category.

Many different types of fully programmable high-pressure leak testing devices are available, such as hydraulic pressure testers and gas leak pressure testers.

Hydraulic pressure testers use distilled water as the pressure medium and deliver up to 2,000 psi. These devices are used for burst testing, fatigue testing and compliance measurement.

Gas pressure leak testers often operate with pure dry nitrogen. They are typically used for final product inspection, because the nitrogen gas does not contaminate the product. In most industries, pressure decay or mass flow using air is the preferred method for leak testing.

Helium mass-spectrometers are often used for testing large volume, non-rigid plastic parts, such as automotive fuel tanks. "Leak rate specifications of 5x10E-4 make pressure decay and other test methods impractical," says Don Carmichael, director of sales and marketing at Vacuum Technology Inc. (Oak Ridge, TN). "In the fully automated test cycle, the fuel tanks are filled with helium at a very low pressure inside a vacuum chamber, and differential-pressure monitoring and control are used to ensure that the tank is not damaged during the evacuation, helium filling and venting portions of the cycle. Sensitive mass-spectrometer monitoring of the vacuum test chamber detects any leakage of helium from the fuel tank."

However, Jacques Hoffmann, president of InterTech Development Co. (Skokie, IL), says "a significant developing trend is checking for leaks in the 10xE-4 to 10xE-6 range with tracer gases that do not require use of a mass spectrometer." His company has developed a system that uses helium in combination with nuclear magnetic resonance technology to both measure and localize a leak.

With pressure decay leak testing, a component is pressurized, isolated from the pressure source, and monitored to see if the pressure decreases. Because air moves from a high-pressure area to a low-pressure area, a leak path can be detected in the chamber. Any reduction in pressure over time is displayed as a pressure change rate. Or, the volumetric leakage rate can be calculated from the pressure reduction.

In the differential pressure leak test, a reference volume is pressurized along with the test part. A transducer reads any pressure differential that develops over time between the leaking part and the nonleaking reference. Differential pressure testing is well-suited to applications requiring relatively high test pressures. And, it typically provides higher sensitivity, more repeatability and faster test times than pressure decay testing.

Pressure decay and mass flow testing are based on the ideal gas law (PV = nRT), where volume and temperature changes affect the pressure in the part as much as the loss of air. "With the reduction in cost of very capable mass flow measurement instruments, this is by far the best approach to leak testing," claims ATS Test Systems' Wilson. "Response times are fast and leak rates are accurate. Pressure decay or differential pressure is now only suggested for very small volume part applications."

Wilson's company is currently developing a high vacuum pressure rise leak testing system. "It appears to be very sensitive for fine leaks, but at a fraction of the cost of traditional helium mass leak testing," notes Wilson. "It is well suited to applications where the absolute leak rate is not critical, just a confirmation that the leak is less than some known rate."

Pressure decay testing is improving in terms of resolution, repeatability, reliability and interfaceability, due to several factors. Although pressure decay and mass flow testing are dependent on the ideal gas law and are therefore limited by basic physics, innovations have improved their performance.

"Highly reliable pressure transducers provide a sensitive and repeatable output for very small changes in pressure," says Gary Grebe, marketing director at Cincinnati Test Systems Inc. (Village of Cleves, OH). "The implementation of the 24 bit A/D converter several years ago allowed for simpler gauge or absolute transducer pneumatics to replace the traditional differential transducer pneumatics. With the 24 bit A/D, pressure changes of 0.0001 to 0.00001psi can be detected by a pneumatic circuit with as little as two valves."

According to Grebe, the pneumatics can be compact, with minimal internal volume. The number of joints in the interconnecting plumbing can be minimized to eliminate potential leak points, thus improving reliability. Improved signal processing of pressure transducer outputs allows for noise analysis and reduction.

"This improves repeatability and test reliability," claims Grebe. "Also, improved test result analysis has allowed for dynamic process tuning to minimize calibration drift due to slow-changing plant or part-to-part temperatures. The evolution of these technological advances has kept pressure decay testing as the simplest and most cost-effective method for production leak testing."

Indeed, many end users are looking for proven technology. For instance, David Kralovetz, vice president of plastic assembly sales at Forward Technology Inc. (Cokata, MN), says he has seen an increase in the use of hand-held mass spectrometer sensors for use in detecting leakage location, even though the technology has been available for more than five years.

"They are often used for testing metal parts, such as automotive radiators, which may be repaired after the leak is located," explains Kralovetz. "Typically, a helium-air mixture is used to fill the part and the sensor detects the location of a helium-air leak."

Kralovetz also says he sees more functional testing, such as performing flow testing through water softener valves and testing valve flow using air at multiple pressure settings to determine flow rate. "Performing functional testing during the leak test allows multiple steps to be incorporated into a single machine and a single basic operation, eliminating the requirement for individual time-consuming operations," claims Kralovetz.

Improving Throughput

To reduce manufacturing cost, the leak testing function must give immediate feedback on specific problems. With feedback, engineers can make corrective actions to the specific areas causing the out-of-tolerance conditions. By reducing scrap, this capability can have a major impact on a company's profit.

Many leak testing devices are harnessing technical advances in transducers, A/D converters, electronics, software and communication tools to put the leak test closer to the appropriate manufacturing process where immediate feedback is important. Network interface via Ethernet, RS485 or RS23 allows for simple interchange of test information with a PLC to the factory computer.

"Effective feedback is more than just pass-fail results that separate parts for shipment from scrap," says Grebe. "Effective feedback is the immediate feedback of leak location. Knowing leak location identifies specific process functions that can be corrected. Immediate process feedback also means getting the leak testing function closer to the operation causing the defect so that corrective action can immediately be taken before multiple defects are produced."

Grebe's company recently unveiled a test system that uses intelligent sensors to detect leak location and total accumulative leak rate. Strategically placed helium-sensitive probes quickly detect the location of a helium leak from the part. After identifying the location, the system measures the accumulative leak rate for the entire product under test. Within a test cycle an accept-reject decision is made and a specific leak location identified to the operator via a computer screen with a part picture.

"The part test records also include the leak location and leak rate information," says Grebe. "At completion of the test, the operator can alert the preceding process of the problem or the test results can be sent upstream in the process. Results are immediate and specific. The equipment is simple and reliable. It is effective for smaller leak rates than pressure decay, mass flow, or vacuum decay or flow. The results are independent of temperature changes."

Many leak testing equipment suppliers are also attempting to improve data collection and interpretation functions. For instance, there is more use of signature analysis tools for reducing test times by analyzing the tell-tale points of a part's pressure decay curve, as opposed to simply monitoring the overall pressure decay of the part over a time period.

Some leak test systems rely on predictive analysis, which often decreases test time while sacrificing reliability and repeatability. Predictive software that looks at the signature of leak test cycles for a number of good parts can be used to predict pass or reject results as early as the fill stage of the leak test cycle.

Math-intensive software is being developed to increase throughput. "Fast sampling rates using 16-bit converters and curve fitting to first order differential equations are allowing us to predict with high accuracy what will be the final leak rate result long before steady state stability in the leak measurement is reached," notes ATS Test Systems' Wilson.

Sciemetric Instruments Inc. (Ottawa, ON) has developed a system that relies on advances in fill and pressure control technology to decrease test cycle time by using closed-loop pressure control to fill and maintain pressure in a cavity. "It uses advanced control algorithms to reduce the time required to fill and stabilize the cavity under test," says Don Parr, leak test product manager. "As a result, the end user achieves reliable and repeatable results in a fraction of the time the previous testing required."

New technology is also being developed to shorten test cycle times. For instance, Grebe says helium-sensitive sensors can detect very small parts-per-million changes in helium concentration outside the test part. "These sensors operate at atmospheric conditions and therefore respond immediately to any leaks when the part is pressurized," says Grebe. "Because the technology does not operate on the ideal gas law, there is no delay time for internal part adiabatic temperature changes or part expansion. Parts do not have to be at or near ambient temperature."

Changes in plant or part temperature, whether fast or slow, do not affect this test technology. Very small leak rates create a fast response at the sensor location. "Cycle times for many parts traditionally tested with pressure decay or mass flow can be significantly reduced because the technology responds immediately to the presence of helium outside the part at the leak location," notes Grebe. "Its test effectiveness is not influenced by temperature, part expansion or virtual leaks."

New Applications

The medical device industry continues to present numerous leak testing opportunities and challenges. Because many medical devices, such as blood filters, have fluid management functions, they require complete leak testing. "Medical and pharmaceutical products present a number of challenging applications for leak testing," says Martin Bryant, marketing manager at Uson LP (Houston). "The medical device market is continually developing new products that have critical functions in which the assurance of leak integrity is vital."

The automotive industry also continues to be a major market for leak testing equipment, with applications ranging from air bags to fuel filters. Recent environmental initiatives, such as the California fuel vapor emission requirements, are forcing automakers and their suppliers to invest in leak testing equipment to ensure that products such as fuel lines, charcoal canisters, gas pump sender flanges and filler tubes are all fuel vapor tight.

Emerging technologies pose new opportunities and challenges for leak testing technology. For instance, fuel cells offer huge potential.

A fuel cell generates electricity through an electrochemical reaction using hydrogen and oxygen. Hydrogen is sent into one side of a proton exchange membrane. The hydrogen proton travels through the membrane, while the electron enters an electrical circuit, creating a DC electrical current. On the other side of the membrane, the proton and electron are recombined and mixed with oxygen from room air, forming pure water and creating electricity and heat. Individual membrane electrode assemblies are stacked in a "sandwich" to boost voltage.

Demand for fuel cells is currently limited, but it is expected to grow dramatically over the next 10 years as the technology is perfected and mass-production processes are refined. According to the Freedonia Group Inc. (Cleveland), the U.S. fuel cell market will skyrocket from $110 million in 2004 to more than $4.6 billion by 2013. PricewaterhouseCoopers (New York) predicts that the global fuel cell industry will reach $46 billion by 2011. Electric power generation applications will drive the market, followed by portable electronics and motor vehicles.

Mike Krumpelt, manager of the fuel cell development program at Argonne National Laboratory (Argonne, IL), says fuel cells contain numerous components that must be thoroughly tested for leaks, such as fuel processors, storage tanks, heat exchangers and interconnects, in addition to intricate tubing. "Hydrogen is a very flammable fuel and it also ignites very easily," warns Krumpelt. "We need to be careful how we engineer products so they are safe."

Krumpelt claims that leakage from the edge of fuel cells is an "order of magnitude greater than from anywhere else." He says the seals at the periphery of the stacks and any connections between system components are most vulnerable.

"The fuel cell industry has always been viewed as the next frontier for new types of applications," notes Cincinnati Test Systems' Grebe. "It offers the potential of high volume because it can relate to the automobile industry, home heating and electricity, and small personal items, as well as the national infrastructure to support it. The timing has always been an issue. If the current energy situation should worsen, it might trigger a national referendum to be energy independent and therefore start change. A major national effort to adopt fuel cells would create many new leak testing application opportunities."

Fuel Cell Challenges

Fuel cells pose unique leak testing challenges, such as:

• Developing appropriate leak specifications.

• Educating potential customers about available leak testing technologies and their limitations.

• Adapting current mass spectrometer technology to hydrogen leak specs.

Typical fuel cell gases, such as hydrogen, are lighter than air and disperse quickly in air. This can increase the time required to trace a leak. The small size of the hydrogen molecule also makes it more difficult to seal against than gases with larger molecules, such as nitrogen or oxygen.

Fuel cell manufactures are attempting to build parts with thousands of potential leak paths and a requirement to be absolutely leak tight. Challenges include very large surface areas with very small internal volumes; parallel leak paths between adjacent flow circuits; difficulty in reliably sealing parts in a production environment; carbon dust given off from the parts; very low pass-fail leak rates; and difficulty in isolating the leak location.

"One of the difficulties that may be encountered is testing the thin, fragile plates," says Uson's Bryant. "Suitable tooling and fixturing is vital. In fact, one of the main technical challenges in all leak testing applications is the ability to make good fixtures to control the position and movement of the part while making a good seal for the part-tester interface."

The fuel delivery system, which includes a maze of valves, regulators, flowmeters, hoses and fittings, is a major area of concern, because there are so many possibilities for a leak. "The fuel cell itself is the other major area of concern, because that's where the hydrogen could come into contact with ignition sources or oxidizers," says Eric Fuller, technical officer for prototyping, integration and evaluation at the NRC Institute for Fuel Cell Innovation (Vancouver, BC). "External leaks, where the gas escapes to the outside world, are relatively easy to detect. Internal leaks, anode-to-cathode, anode (or cathode)-to-coolant, are inferred by measuring other effects."

According to Fuller, fuel cells require several different types of leak testing. "A typical leak test involves a soap and water mixture which is dripped onto suspected leak areas," he explains. "This liquid forms bubbles if gas is leaking in that area. This is a simple and inexpensive test that can be used for all gases. However, this solution is typically conductive and therefore it cannot be used directly on fuel cells."

Another common leak test involves sealing all the inlet and outlet ports on the fuel cell, then pressurizing the fuel cell with an inert gas. Once the fuel cell is pressurized, the gas supply is disconnected, and the pressure of the fuel cell is measured. If there is a leak, the pressure will drop. "This is a good way to measure internal leaks on a fuel cell or heat exchanger," claims Fuller.

Another test method, which is used for very small leaks, involves pressurizing the fuel cell with helium. A gas detector, such as a mass spectrometer, is used to detect the helium as it leaks out of the fuel cell. "An inverted version of this test involves evacuating the fuel cell through a mass spectrometer, then spraying helium onto the suspected leak areas," says Fuller. "If there is a leak, the helium will find its way into the mass spectrometer and be detected. This method can be very time consuming if the fuel cell is very large."