Adhesive bonding technologies for thermoset polymer composites have been used in marine, automotive, construction and aerospace industries due to their superior mechanical behaviour (high strength-to-weight ratio, damage tolerance and fatigue resistance) compared to conventional joining methods. The main disadvantage of this joining technology is its susceptibility to delamination due to disbonding during use. Loading conditions, adhesive type, ageing effects and lack of inspection procedures are just some of the elements that affect the overall structural performance of the composite joint during the manufacturing process. A deeper understanding of how these elements affect joint behaviour is required to improve joint performance and design. This work provides a comparative fractographic analysis for two different joint types: co-bonded (CB) and secondary bonded (SB) joints, under Mode I delamination at elevated temperature and high humidity conditions. Fractographic analysis was used to compare the two joint technologies and explain the differences in toughness values and fracture behaviour, revealing crack propagation mechanisms in composite joints. While the CB and SB joints have comparable fracture toughness (G_IC) values, different fracture characteristics and bonding methods can discern these two bonding technologies, indicating that SB joints are more susceptible to environmental conditioning.

Our aim with this work was to evaluate how the thermoplastic resin used in the composite adherends and on the energy director affected the static ultrasonic welding process in both parallel and misaligned configurations. Polyetheretherketone (PEEK) and low-melt polyaryletherketone (LMPAEK) were the resins used and their thermomechanical properties were characterized via dynamic-mechanical analysis and modulated differential scanning calorimetry. With parallel adherends, neither the welding time nor the through-thickness heating in the adherends vary significantly. This similarity was attributed to a larger heat capacity of the PEEK energy director counterbalancing its higher viscoelastic heating rate. With misaligned adherends, the welding time was larger for PEEK welds than for LMPAEK welds and LMPAEK adherends presented a larger though-thickness heating. These effects were attributed to the larger bulk viscoelastic heating rate of carbon fibre reinforced/LMPAEK adherends adding up to the lower heat capacity of LMPAEK.

This paper evaluates the potential of changing the welding force and the compliance of the energy director (ED) to reduce the effects caused by misaligned adherends, which were: increased through-thickness heating, reduced size of welded area and increased heating time. In the methodology that was followed, we welded adherends misaligned by approximately 4.5° in different scenarios: with higher welding force; with increased ED compliance by the use of a thicker ED and; with increased ED compliance by the use of a discontinuous ED. The most significant reduction of the effects caused by misaligned adherends was obtained when combining the use of both increased welding force and discontinuous ED. Such improvement derives from the imposed parallelism caused by the use of a higher welding force and from a more efficient concentration of heat generation at the weld line that occurs when a discontinuous ED is used.