Leak Test Technology Development (LTTD) & Prove-out
DoD Participants: U.S. Navy – Fleet Readiness Center Southeast
Current state-of-the-art leak detection in industrial and military production communities uses helium mass spectrometry to detect leaks in products or components with sensitivity capability up to 10-11 standard cubic centimeters per second (sccs). However, sensitivity levels predominantly sought in industrial and military production applications range from 10-2 to 10-6 sccs.
Typically, vacuum–helium mass spectrometry is used on-line to detect and quantify cumulative leak rates within this range. Then, a second stage operation at a reject station is needed to isolate and locate the leak site or sites on the product. Once at the reject station, the part is typically repressurized to the requisite psi and subjected to testing by immersion in water or other liquid in a dunk tank. This process has sensitivity in the 10-2 sccs range leaks. A single leak site having a leak rate volume at 2 x 10-6 sccs would take over 135 minutes in a dunk tank to form one bubble. Larger products, where total immersion is not practical, are typically subjected to pressurization/soaping for leak location. This bubble detection methodology has all the same limitations as does total immersion, low sensitivity and total dependence upon operator diligence and acuity.
Today’s process results in either many parts being scrapped because the leak location cannot be found at the reject station and marked for repair, or by low detection sensitivity methods, like immersion and soaping, which allow many parts with latent leaks to pass into final assembly.
In January 1999, Leak Test Prototype Development & Prove-Out was funded jointly by the Office of Secretary of Defense (OSD) and the industrial partners in the National Center for Manufacturing Sciences (NCMS) Leak Test Technology Development (LTTD) Consortium under Cooperative Agreement DASW01-98-2-0002. Prototype technologies were built and tested for NCMS member industrial applications and for defense depot applications at Fleet Readiness Center Southeast.
Three technologies advanced out of research done earlier by the LTTD Consortium: speed of sound (SoS) sensor, a derivative of surface acoustic wave (SAW) research work, photoacoustics (PA), and millimeter wave (MMW). The SoS sensor and MMW were targeted for prove-out on applications at Fleet Readiness Center Southeast in addition to industrial applications. Argonne National Laboratory (ANL) was to build the prototypes with assistance of Vacuum Instrument Corporation (VIC). A PA prototype was to be built by VIC for targeted applications in the ground transportation industry and weapons manufacturing at Honeywell Federal Manufacturing & Technologies, the Department of Energy’s Kansas City Plant (KCP).
The SoS sensor uses two transducers, a gas pump, and a single- or dual-piezoelectric cavity in which sound waves propagate. Its simplicity results in a sensor that is a robust and economical method for detecting leaks of helium, hydrogen, sulfur hexafluoride (SF6) and other gases having molecular weights lighter or heavier than ambient air.
The SoS leak detector has the following practical advantages and features:
· Relatively high sensitivity (≈10-4 sccs)
· Fast response time, approximately seven readings per second
· Reconfigurable zero offset to compensate for ambient tracer gas concentration changes
· Detection of very large leaks without saturation
· No heating components
· Vacuum not required
· Calibration not required
· Detection of lighter gases (hydrogen, helium) and heavier gases (freon, SF6)
· Detection of combustible gases, such as fuel vapor, methane, and hydrogen
· Rugged sensor design with no moving parts other than an air pump — minimal maintenance
· Lightweight (< 5 lb.), portable unit can be worn on a person
· AD/DC adapter and rechargeable battery
· Audio and visual user interfaces
· Multiple outputs — audio, visual, light emitting diode (LED) display, and RS232
· Economical cost
· Is intrinsically safe.
The field test of a single-cavity SoS sensor prototype was performed at Fleet Readiness Center Southeast on F-14 fuel system lines. The primary detection means for leak localization, once a leak had been detected by pressure decay methods, was to use a surfactant (soaping) and look for bubbles forming. While this method is simple to execute in exposed areas such as wing surfaces, it is not very sensitive and not all parts of the fuel system are accessible and shaped in such a manner that a surfactant can be applied and observed for leakage.
The tracer gas helium, mixed with oxygen and supplied by a charge station built for Fleet Readiness Center Southeast, is detected by the handheld portable SoS sensor with a pass/fail display easily viewable by the user. The advantage over conventional means of leak detection was this helium leak test system could detect leaks 10 times smaller than the leak testing methods being used at Fleet Readiness Center Southeast. It is safe and it utilizes a single test apparatus to perform a multitude of leak test procedures. Pressure decay tests can be performed manually or automatically on the station to detect gross leaks using shop air before injecting the helium mixture into the fuel lines. The SoS sensor, with its increased localization and sensitivity, and the charge station have provided Fleet Readiness Center Southeast with a state-of-the-art leak detection system that is rugged enough for field use and inexpensive enough to proliferate widely throughout Navy depot and field operations.
The F-14 aircraft is a high performance United States (U.S.) Navy fighter jet that flies in a very harsh environment. Because of this harsh environment, the F-14 aircraft is taken out of service approximately every seven years for Standard Depot Level Maintenance (SDLM) which is a complete tear down and inspection of its internal components and structure.
The reassembly phase of the F-14 SDLM effort requires leak testing of multiple tubes within the F-14 fuel system. Finding and resolving leaks is essential to the proper function of the fuel system. The first prototype test was performed on an F-14 aircraft in August 2000.
A variable-length hollow wand allows the lightweight, portable SoS to probe and sample hard-to-get-to areas while strapped over the technician’s shoulder. Helium concentrations were read off an LED on the top of the sensor, or a threshold could be set and when exceeded, would provide an audible tone through earphones and/ or visually with a head-mounted warning light.
While pressurized with helium, the SoS detector was used to “sniff” all the fuel system joints, fittings and component interfaces within the eight F-14 fuselage fuel cells and engine nacelle. Numerous leak paths were quickly identified and, interestingly enough, one verified leak was not detectable using the bubble method or the pressure decay method. The undetected leak was very significant, as it was located in the engine nacelle, an area that requires high-maintenance hours to access after aircraft buildup and has many high-temperature ignition sources. Maintenance man-hours and safety are two major concerns for the Navy fleet and are crucial to the ability of the F-14 aircraft to keep flying and flying safely. Four additional F-14 aircraft have been tested using the helium detection method with similar results. This truly makes the helium detection system a significant improvement over the outdated current leak detection methods.
Because of the success on the F-14 aircraft, the F-14 Fleet Support Team has offered this system to other areas at Fleet Readiness Center Southeast with positive responses.
Ford Motor Company used the SoS sensor to find the leak sites in the fuel cell manufacturing process. The fuel cell uses a stacked series of plates with manifolds for hydrogen gas and oxygen from air to generate electrical power. These plates are stacked together to a particular height to produce the required voltage level. Defects in the seals and plate manufacturing process can lead to leaks in the internal manifold that could not be localized before utilizing the SoS sensor.
VIC in association with the oil drilling industry, arranged for testing of the SoS sensor in oil field drilling operations. The SoS was tested in head-to-head trials with a commercially available helium detector. The industry currently uses a sniffer-based mass spectrometer, which poses many operating problems in the field environment. These trials resulted in the SoS being selected as the leak detector of choice by those performing the oil field trials.
PA technology uses an infrared laser that is scanned across a part containing the tracer gas, SF6. Leaking SF6 gas strongly absorbs the laser light and undergoes a rapid expansion giving an acoustical emission, which is detected by an array of microphones.
The advantages of the PA technology are:
· High sensitivity (~10-6 sccs SF6)
· Low capital costs compared to helium mass spectrometers
· Reduced cycle time (~1:3 ratio in cycle time/facilities requirements)
· Eliminated the maintenance issues associated with large vacuum pumps and air tight enclosures
· Determination of leak location to within 1 mm.
This PA prototype was designed as a manually-loaded single chamber system. Because this prototype was a first of its kind, it was not feasible to pilot this prototype in automated production critical-path areas. For this reason, the PA prototype was designed to operate in an off-line location such as the repair area during the trial phase. As a repair area leak test station, the PA prototype was intended to recheck rejected parts coming from the mainline leak test system.
The test parts for which the system was designed included an automobile climate-control system compressor, a drive-train torque converter, and an undisclosed classified application at KCP.
For a 1 x 10-5 sccs SF6 calibrated leak, the PA localization system developed by the LTTD Consortium gave excellent sensitivity and position detection results under noisy factory conditions. Tests with a smaller 1 x 10-6 sccs leak showed only marginal sensitivity detection and poor position detection. High ambient noise in the factory while conducting the test runs without an acoustical chamber interfered with performance of the system on the 10-6 sccs leak size. The LTTD Consortium identified a series of design and engineering changes to improve position detection at the 10-6 sccs leak level using an acoustic chamber to minimize noise.
The scientific principle uses the property of microwave/MMW rotational absorption of leak gases. MMW spectroscopic techniques offer great potential for standoff detection of polar gases. Selecting an appropriate gas for its absorption properties is critical to the success of this technique and was, as it turned out, the Achilles’ heel for this prototype.
Out of a list of 41 polar gases submitted to the LTTD Consortium by ANL, 2 gases, nitrous oxide (N2O) and Freon 134A (C2H2F4), were finally selected for MMW tests in industrial applications based on their spectral properties and safety compliance for the applications under consideration (engine block and torque converter). Neither of these gases was judged suitable for use in the Fleet Readiness Center Southeast application, the F-14 fuel bladders. For Fleet Readiness Center Southeast, the LTTD Consortium examined the feasibility of building a standoff microwave/MMW sensor for detection and location of aircraft fuel bladder leaks using ammonia as the tracer gas.
Ammonia, currently used in the testing done at Fleet Readiness Center Southeast, has strong absorption at ~24 GHz. The plan was to build a microwave sensor with an x–y scanner used to scan the MMW beam over the inflated bladder and the reflected or backscattered signal would be detected by a diode detector. The signal change due to absorption of the MMW energy by the leaking ammonia gas would be measured to identify leaks. A special signal processing scheme would be developed to discriminate the leak signal from the background fluctuations caused by the part geometry.
Ammonia, nitrous oxide and Freon 134A tracer gases were examined using two different configurations of MMW. Two different MMW apparatus were built and bench tested: one using a vacuum chamber in an attempt to heighten sensitivity for the industrial applications, and the second using an open-space swept-frequency technique for the Fleet Readiness Center Southeast application.
After several bench tests and equipment modifications it was finally concluded that neither method provided the results sought. The swept-frequency method was found to yield low sensitivity due in part to diffusion and humidity; both would be present in the open-space environment of the test area at Fleet Readiness Center Southeast. As a result, further work on this prototype was terminated by the LTTD Consortium with consensus from Fleet Readiness Center Southeast and KCP.
An invention disclosure was made on innovations developed in this technology by ANL.
The SoS sensor would benefit from additional work in temperature compensation, investigation of its acoustical properties at low pressures for the National Aeronautics and Space Administration (NASA) space plane application, and evaluation of sensitivity improvements that could be achieved in the dual-cavity design.
Future work on the PA localization system as an on-line prototype should focus on acoustical signal strengthening as it relates to part geometry, chamber reflective and acoustic absorption surfaces, and developing a robust quantification routine of the acoustic signal.
The LTTD Consortium hopes to launch, along with four DoD repair depots, a follow-on collaborative program to investigate further the benefits and quality improvements that can be brought by these and other newly emerging leak testing technologies to a wide variety of industrial and military production and repair applications. In addition to realizing the benefits to be derived by the user communities, the knowledge gained in these pilot tests will allow the commercializing companies in the Consortium to harden designs and speed them to market acceptance.
Four new technologies are currently being evaluated for such a project: two developed in the LTTD project just completed and two developed totally at private expense. They are:
1. PA (see discussion above)
2. SoS sensor (see discussion above)
3. High Definition GasVue developed by Laser Imaging Systems (LIS)
4. USAS detector, developed by VIC.
Sensitivity capability will range from 10-4 to 10-11 sccs with these four unique technologies.
A portable helium leak detector has been successfully developed and demonstrated based on single-cavity measurement of SoS variations in helium–air mixtures. The SoS technology is suitable for any application where a standard gas is present and traces of other molecular weight gases exist that represent some anomaly, leak pathway, or volumetric contamination.[1]
Four prototypes were built and tested in Phase II, each containing incremental improvements.
Prototype IV has leak rate sensitivity of 2 x 10-4 sccs, our most sensitive sensor to date.
U.S. Patent 6,279,378, Ultrasonic Gas Analyzer and Method to Analyze Trace Gases for the device and method was issued to the University of Chicago, operator of ANL on August 28, 2001. Each industrial member of the LTTD Consortium holds a non-exclusive license to the SoS technology developed in Phase I and II through NCMS. VIC began commercialization of the SoS upon conclusion of the Phase II project.
The Phase II PA localization prototype stands alone as the only technology capable of detecting both presence and location for leaks as small as 1 x 10-5 to 10-6 sccs SF6. PA, a point inspection method, measures each leak site individually but is capable of locating multiple leak sites on a single part.
PA also has unique potential for batch applications. MFP can identify the location of one or more leaking parts in the same cycle.
Commercialization of the PA localization system is expected to proceed under VIC with PA laser scanning detection sub-system available from LIS. NCMS and Ford Motor Company retain sublicensing capability to the PA localization and SoS technologies to ensure future commercialization.
SoS Sensor Process
· The SoS system is portable, easy to use and inexpensive.
· Field trials support system’s ruggedness and durability.
· System has excellent user interface with sound and optical feedback ensuring operator attention to leak resolution.
· Can detect leaks 10 times smaller than current leak testing methods used in F-14 SDLM Fleet Readiness Center Southeast operations. Found leak in engine nacelle not able to be found via Fleet Readiness Center Southeast methods in use today
· Individual leaks can be detected in under 1 sec; unit recovers immediately.
· Outperformed a helium sniffer-based mass spectrometer in adverse oil field conditions.
· Throughput may remain the same, but mean time between failures (fuel leaks in the field) should increase significantly.
· Costs of lost fuel, labor costs associated with fuel evacuation and repair will be significantly reduced as mean time between failures increase.
· Reductions in failures at the wet fuel check stage will be realized.
· Tracer gas is formulated to be a breathable mixture of helium and oxygen which eliminates the asphyxiation danger posed by use of 100% nitrogen in confined space fuel line checking done currently.
· Return on investment in the first year is over 100%. Cost of one engine pull covers the cost of one SoS sensor.
· Method can detect and locate multiple leak sites in a 2-ft2 area in 6 sec cycle time.
· Reduces the number of stations required roughly by half translating to headcount and facility savings by reducing floorspace and line size.
· Part leak sensitivity down to the 10-6 sccs range is possible.
· Automatically determines the leak site and plots the leak on a photograph of the part showing where the leak is situated thereby eliminating a second stage operation.
· One PA station can support production throughput equal to that supported by two mass spectrometer stations currently.
· Provides immediate feedback on leak location.
· More efficient than traditional methods, thus reducing employed man-hours.
· Repeat service visits will be dramatically reduced, increasing customer satisfaction.
· Ford’s McBean Analyses calculated an average R/I with facilities costs at 1.0 for one unit in first year and 8.0 for three units in the fourth year.
[1] Shuh-Haw Sheen, Haul-Te Chien and Apostolos C. Raptis, “Ultrasonic Techniques for Detecting Helium Leaks,” Sensors and Actuators, Vol. B71, No. 3, December 2002, pp. 197-202.