TWI Industrial Member Report Summary 929/2009
Decourcelle N and Kellar E J C
Background
The use of adhesive bonding has increased during the past few years in many industry sectors. A number of advantages have made it significantly attractive to use, such as the ability to join dissimilar materials, more flexibility in product design and improved in corrosion resistance, and stress distribution in the joint. However, where high reliability is required, it is used very prudently due to limitations, such as the mechanical properties (peel and shear), the necessity for surface pre-treatment, temperature limitations, control of process parameters, and restrictions in inspection methods for bond quality.
Many non-destructive methods have been investigated to detect defects within adhesive bonds and to determine the reliability of bonded joints (Munns and Georgiou, 1994; Adams and Cawley, 1988). However, there is not one single technique able to guarantee the detection of all defect types found in an adhesive bond. In addition, there is one defect referred to as a kissing bond which cannot to-date be reliably detected by any conventional non-destructive testing (NDT) technique. The term 'kissing bond' should refer to a condition in adhesive bonding where there is intimate contact between the adhesive and the structure/component but without adhesion or molecular bonding between these two entities. In fact, a joint containing a kissing bond is not totally disbonded, it is identical to a perfect joint but possesses lower adhesion strength (80-90% reduction), and is for this reason classified as a critical defect. This form of defect has proved difficult to simulate but it represents one of the main limiting factors against the widescale use of adhesive bonding in industries such as aerospace.
A literature review previously conducted (Decourcelle, 2007) showed that many techniques for adhesive joint inspection exist with different abilities and limitations. It is clear that none of them has yet been proven to detect and characterise kissing bonds, and research is still in progress on this subject particularly in the civil aviation industry. Many previous studies (Tavrou et al, 2002; Freemantle and Challis, 1997; Smith et al, 2003) have been carried out to develop normal incidence ultrasonic longitudinal wave techniques with variable levels of success, using different forms of signal processing such as spectral analysis, signal attenuation, signal absorption, resonance, etc. They all came up with the same findings that these techniques are only sensitive to volumetric defects and disbonds at the interfaces. Nevertheless, Brotherhood et al (2003) carried out ultrasonic inspections on adhesive joints using conventional normal incidence longitudinal and shear waves with a standard immersion transducer and an electro-magnetic acoustic transducer (EMAT). The aim of this study was to assess the detectability of kissing bonds by calculating the reflection coefficient of the interface at varying loads for a number of surface roughnesses. The results showed a good sensitivity for the techniques involved, but the main unknown factor was how the kissing bonds had been produced in the samples and whether or not they really were realistic. Another former body of work (Pilarski and Rose, 1988) studied the ability of an ultrasonic shear wave technique using oblique incidence (30° to the adhesive layer) to detect poor surface preparation in a bonded joint of two aluminium plates. The results revealed that this technique possessed good sensitivity for the detection of imperfections appearing at the adhesive/adherend interface, and showed that shear waves are more capable of detecting these interfacial problems than normal incident longitudinal waves. Therefore it appears that a technique using ultrasonic shear waves could be a good candidate in the detection of kissing bonds seeing that these defects are present at the adhesive/adherend interface. For this reason, this feasibility study focused on the utilisation of two techniques involving ultrasonic shear waves, a through-transmission and a pitch-catch technique.
Objectives
- Develop a consistent and reproducible methodology for producing satisfactory sets of samples meeting the kissing bond requirements as defined in the literature.
- Identify ultrasonic NDT techniques suitable for finding adhesive kissing bond flaws.