Quality control adhesive-bonded joints, defects

For the processing quality control of adhesively bonded joints, scientific approach based on both probabilistic design and statistical treatments are necessary. For deceasing the ratio of defective joints, the coefficient of variation (CV) of the samples should be kept small. The cohesive fracture ratio (C/T) of fractured joint surfaces is vitally important to keep CV small. Since the cohesive fi"acture ratio decreases often due to mechanical or chemical mismatch of adherends and adhesives, the surface treatment of adherends is important. 

Proper intervention planning and design, workmanship and quality control are essential since the non-reversibility of adhesion does not allow inspection and correction after cure. Compared to currently used mechanical timber joints, bonding requires more skilled and well-trained operators, and better organisation and record keeping for future traceability of possible problems and defective joints. 

A different principle is used in ultrasonic resonance testing. Here, thickness vibrations are induced in the adherend/adhesive/adherend sandwich in which the two adherends eire considered to act as rigid masses while the adhesive is an almost massless spring. If the type and thickness of the metal sheets is constant, then the resonant frequency is a function of the thickness and bulk modulus of the adhesive. The well-known Fokker Bond Tester (Mark II) overcomes the difficulty of the many variables in the ultrasonic resonance test by coupling the joint to a system of well-defined resonance characteristics. Changes in the frequencies of the joint-transducer system are correlated with known defects and a calibration curve obtained. The calibration curves can then be used to establish acceptance limits which can be (and are extensively) used for quality control in production. 

---