Simulated aircraft structural joints using CFRP adherends and a supported (for bond thickness control) and an unsupported version of a 120°C epoxy adhesive were fatigued at five different combinations of temperature and humidity. These combinations were: ambient dry, and at 85%RH, 90°C dry, and at 65%RH, and -50°C. The high humidity fatiguing made use of a saturated salt humidity chamber specially developed for the purpose.
The joint geometry used was a variation of a cracked-lap shear joint chosen because it simulated actual joints, allowed the progress of fatigue damage to be observed, and was simple to manufacture.
Fatigue damage was monitored and assessed via a combination of microscopy, ultrasonic inspection and radiography.
The results obtained showed that the fatigue behaviour of adhesive joints was considerably affected by temperature, humidity, and support medium. Cohesive failure of adhesive, interfacial failure, and interlaminar failure of adherends were all observed, such that predictions of structural integrity would be difficult to make. In addition, failure modes observed could be aggravated by, or themselves exaggerate, gross defects or inconsistencies already present in a structure, e.g. impact damage. Optimum static performance may not give acceptable fatigue behaviour.