Ultrasonic Bond Testing as a Quality Tool
With the growth in the use of adhesive bonding and composites, bond testing is seeing its second wind.
The use of adhesive bonding to join metals, composites, plastics and other materials is increasing across many industries. The aircraft industry widely adopted metal to metal bonding after World War II and this has since spread to different material combinations and other industrial sectors including modern aerospace, automotive, marine and wind energy.
Bonded joints offer many advantages over conventional methods such as bolting, riveting and welding. Weight saving is key but other benefits include an improved distribution of stresses by eliminating the need for holes and welds, easier design and assembly of large area joints or curved external surfaces and an integrally sealed joint.
The integrity of these bonds is critical to the quality of the final product and so techniques to assess the quality of the bond have seen a resurgence. With new material combinations come new possible flaw types, and the challenge is to develop inspection techniques to detect these.
With such a range of materials and configurations used in multi-layered bonded structures, different types of testing methods have been developed to attempt to cover all combinations.
To view the chart for Bond Testing click here.
Modern bond testing is really an electronic form of one of the oldest inspection methods called “tap testing.” Tap testing in its simplest form requires the inspector to tap the structure with a small hammer, or another object such as a coin, and listen to the sound radiated from the structure. The characteristics of the impact are affected by the impedance of the material and the type of hammer used. Any flaw returns a ‘flat’ or ‘dead’ response when compared to a good area of the structure. The interpretation relies on the subjective tonal discrimination of the operator, requiring substantial experience. More modern electronic tap hammers provide a digital readout that can be correlated with flaw types. The acoustic response of a good structure can vary with geometry and only shallow flaws can be found, typically less than 0.080 inch (2mm) from the surface which limits its use. Sandwich constructions, with two skins separated by a core, therefore require access from both sides of the structure for complete inspection, which is not always possible.
Conventional ultrasonic flaw detectors work by transmitting ultrasonic sound waves into a material which is then reflected or transmitted by a boundary with a different material. Flaws or discontinuities reflect ultrasonic waves and by monitoring the echoes, hidden flaws can be detected. Ultrasonic flaw detection typically utilizes frequencies from 500kHz to 10MHz with lower frequencies providing higher penetration into materials but having less sensitivity to smaller defects because of the longer wavelengths. Multi-layered bonded structures and core materials can be a challenge for conventional ultrasonic testing. Bondlines and core materials may completely dampen the sound, provide multiple echoes which are hard to resolve and therefore preventing inspection. Also, some bonded structures can be damaged by wet coupling agents that are a requirement for conventional ultrasonic inspection, making the technique non-viable.
Modern ultrasonic bond testers are small, lightweight and portable for use in the factory or in-field. They operate at a lower frequency compared to conventional ultrasonic testing, typically between 4kHz-400kHz, enabling deeper penetration in attenuating materials, across multiple glue lines and even across sandwich cores to detect far-side flaws. The microprocessor based instruments have multiple display modes that are optimized for different applications with a variety of gates and alarms to easily identify a flaw.
To cover such a wide variety of applications, modern bond testers were developed to bring together, into one instrument, three different testing modes that, historically, had been developed separately, each having their own preferred applications. Pitch-catch, mechanical impedance analysis (MIA) and resonance testing modes all have some similarities and, like tap testing, the measurement metric is a change in the amplitude and/or phase of the sound transmitted into the material and that the result is a relative measurement between a good and flawed area of structure. Pitch-catch and MIA can be dry-coupled, which is a major advantage when the use of couplant is not possible.
The three modes of operation work as follows:
1. Pitch-Catch mode (10kHz-40kHz):The probe consists of a transmitter and receiver element on separate tips. The transmitter ‘pitches’ a burst of acoustic energy that propagates into the test part, the receiver ‘catches’ the sound. In a bonded condition, the sound waves propagate across the skin (as a plate wave) with some attenuation into the core. With a disbond, there is little attenuation possible into the core giving higher amplitude at the receiver. Pitch-catch can operate with a single frequency burst or as a swept frequency where the burst of acoustic energy is transmitted to the part across a pre-defined swept frequency range. There is also a high-energy pulsed mode, for thicker materials, which transmits a spike pulse of broadband acoustic energy into the part and measures the amplitude only of the received signal. This mode is fast, dry-coupled, easy to calibrate and has high penetration but is limited to defects >0.5 inches (12.7mm) because of the tip spacing on the probe. Far-side flaws can also be readily detected with honeycomb cores acting as waveguides to efficiently transmit sound across the core with more sound reflected where a far side flaw is present.
2. Mechanical Impedance Analysis (4kHz-12kHz):The probe consists of a driver and receiver element coupled in series with a single probe tip. The loading of the receiver element in the probe tip is related to the stiffness of the test part. In the setup process, the optimum test frequency is automatically chosen. When the probe is referenced or ‘nulled’ on a good area, the driver and receiver elements vibrate together at the same phase and amplitude. As the receiving element is moved to a flawed area, which is weaker, the phase and amplitude of the signals change. The loading on the tip is very high over bonded regions and much lower over unbonded regions. Since the measurement is a comparison of stiffness results are better on stiffer structures such as metal skin to metal honeycomb. Flexible composites do not show much change in stiffness between the bonded and unbonded regions.
MIA can operate with a single audible frequency burst or as a swept frequency where the burst of acoustic energy is transmitted to the part across a pre-defined swept frequency range. This mode is dry-coupled, can more accurately locate a flaw because of the single tip and works well on curved or irregular surfaces. Penetration is limited to about 0.1 inch (2.5mm) depth with most applications limited to skin to core flaws in honeycomb structures.
3. Resonance (110kHz-370kHz):Resonance probes are ultrasonic contact probes driven at their resonance frequency. The resonance frequency is determined automatically and on contact with the surface, the probe is dampened reducing the signal amplitude and changing the resonance frequency. By moving from a good to a bad area, the acoustic impedance of the material changes where the structure is weaker and therefore the phase and amplitude of the signal changes. The material acts like a thin plate and a standing wave is set up under the probe. Changes in the effective thickness caused by a flaw will affect the amplitude and phase of the standing wave. The phase change is related to the depth of the defect in multi-layered structures. This mode requires couplant and is best for flat surfaced laminates where it has high penetration and can provide depth information.
All bond testing is a comparative technique and requires good reference standards with known defects to ensure the best equipment set-up parameters, optimum sensitivity and hence the best results. To keep material properties constant, reference standards must be made with the same material as the structure that is to be inspected and have representative flaws which can be a challenge to manufacture. Material properties and flaw types can be very frequency sensitive and so a good calibration on the reference standard is critical. However, with proper calibration, flaws including disbonds, unbonds, foreign objects, delaminations, impact damage, crushed or damaged core, voids, areas of porosity and more can be detected.
Displays reflect the comparative nature of the measurement and show the amplitude and phase shift of the signal compared to the referenced signal from a good area. Results can be displayed in different modes including Radio Frequency (RF) envelope or impedance plane display. The impedance-plane display is a polar coordinate system showing the phase shift and amplitude of the area under inspection compared to the referenced area. In fixed frequency a single point is displayed, hence the term ‘flying dot,’ but in swept mode, a circular display can be plotted representing multiple data points as the frequency is swept between defined limits. Amplitude changes are indicated by a radial distance change of the dot from the center point of the display (null reference point); phase changes are represented by a rotation of the dot around the null reference point. Flaws are normally represented by an increase in amplitude and/or a phase shift. A time-encoded profile of phase and/or amplitude can also be used for rapid scanning which is particularly useful in pitch-catch mode.
Mode Selection Guidelines
With an almost unlimited combination of materials, adhesives and their resulting material properties, determining which the most appropriate testing mode is can be a daunting task. The value of a representative reference standard cannot be understated, but below is a guide to indicate which mode is likely to work most effectively for some common material combinations, requirements and flaw types. The ultimate success of an inspection mode and probability of detection of the flaw depends on the material properties, skin thickness, core thickness, geometry, flaw type, size and depth.
Bond testing is already established as a requirement in maintenance manuals and technical directives for multiple commercial and private aircraft types but is now seeing its use expanding. With such a wide range of material combinations used in multi-layered structures, the need to inspect the integrity of the bonding is critical. New types of flaws, unique to bonded structures, whether produced in manufacturing or as result of in-service use must be detectable and this requires different techniques to conventional NDT inspection. Bond testing offers a solution for many bonded and composite structures and is portable, cost-effective and easy to use. The development of representative reference standards will always be a challenge in manufacturing, but is essential for the most accurate and reliable inspections.