The influence of inclusions on the strength of adhesive joints

Abstract Brittle adhesives based on cyanate ester monomer bisphenol A dicyanate resin have been physically modified in an attempt to improve their fracture toughness and the strength of bonded joints. This used a combination of woven and knitted fabrics based on PEEK monofilament and polyaramid fibre incorporated into the adhesive. Mode I fracture toughness measurements using a peel, double and single cantilever beam test showed increases from ∼60 J/m2 for the neat resin to values of ∼2000 J/m2 for the modified systems, as a result of extensive monofilament and fibre crack bridging. However, there was also an accompanying reduction in lap-shear strength for specimens modified with woven PEEK fabrics as a result of the formation of sub-micron-sized cracks around these monofilaments with their axis parallel to the length of the specimen. These interfacial cracks were observed in-situ on the optical microscope as shadow bands adjacent to the monofilament edges. The shadow bands were formed as a result of total internal reflection at angles of incidence greater than the critical angle. For gaps with widths less than the wavelength of light, it is shown that reflection is not total, and instead some transmittance occurs via frustrated total internal reflection. The distribution of light as a function of incidence angle is calculated for interfacial gaps with widths of λ / 100 , λ / 10 and λ / 2 . It was found that using knitted fabrics, as opposed to woven fabrics, did not result in a reduction in lap-shear strength. This was probably because of the prevention of the monofilament interfacial cracks running along the specimen length.