Development of Infrared Techniques for Practical Defect Identification in Bonded Joints

Abstract

Identification of kissing defects in adhesive bonds has been reported to be an area of concern across a range of industries. To date the majority of work on this matter has focused on the development of advanced ultrasonic techniques. The current thesis focuses on the use of thermography, specifically pulsed and pulse phase thermography (PT and PPT), for the identification of kissing defects. Initially the thesis focuses on the application of PT and PPT for the identification of a range of defect types in a variety of materials to establish the effect of material properties on identification of defects. A numerical model has been developed to simulate the thermal evolution created during a PT or PPT experiment. After validation through a series of case studies, this model has then been used as a predictive tool to relate defect detectability to the thermal property contrast between defect and bulk materials. Where insufficient thermal property contrast exists defects have a limited effect on heat propagation through a component and therefore are not detected using PT or PPT. A means of producing realistic kissing defects in bonded joints is established. The addition of a small load to bonds containing kissing defects was found to open the defects sufficiently to enable their detection. Initial experiments use the application of a tensile load, via a test machine, to successfully investigate simulated kissing defects in single lap joints. A technique using vacuum loading on one adherend of an adhesive bond while PPT is carried out from the other adherend was successfully trialled. Vacuum loading enables the technique to be taken out of the laboratory. A low cost infrared detector, Flir Tau320, compared to the research based photon detector, Flir SC5000, was demonstrated to be suitable for application in PT, thus enabling a significantly lower cost tool to be developed.