Solid rivets are widely used as mechanical fasteners in airframe applications due to their relative low cost and good fatigue performance. Although rivet installation is known to influence this fatigue performance, variabilities in hand riveting practices make exploiting rivet installation as a design variable difficult. Developments in riveting technology have led to force-controlled rivet squeezers and fully automated riveting gantries, improving the consistency of rivet installation and providing the opportunity to exploit its influence on
fatigue.
This dissertation describes a research program undertaken to examine the influence of rivet installation on the fatigue performance of riveted lap joints and identify what aspects can be exploited during design. A combination of finite element analyses and experimental techniques were used to investigate the role of rivet installation on the formation of residual stresses and on secondary bending stresses in a loaded joint, two aspects established as critical to the fatigue performance of riveted lap joints. Crack growth reconstructions of fracture surfaces marked using a special marker fatigue spectrum were also completed in order to quantify the effects of these residual and secondary bending stresses on fatigue performance. Additionally, variations in the effects of rivet installation on traditional monolithic aluminum sheet materials and hybrid aluminum-fibre glass laminates known as GLARE were also investigated.
Results from these investigations provided new insights into the role of rivet installation on fatigue. The radial expansion mechanism to which residual stress formation during riveting is typically attributed was observed to be a secondary mechanism relative to the through-thickness compression of the joined sheets. The location and magnitude of peak secondary bending stresses were found to be directly influenced by rivet head geometry. In
certain cases, shifts in the location of peak secondary bending stress were found to negate the benefits of residual stresses formed during riveting. Finally, differences in the influences of rivet installation on GLARE and monolithic aluminum sheet were also identified. The presence of the fibre layer in GLARE alters the residual and secondary bending stress states, indicating a potential difference in joint designs optimized for aluminum and GLARE.
GLARE (GLAss REinforced) laminates are part of a family of metal-composite hybrid materials being used for airframe applications. One critical design element in such an application is the riveted lap joint. Machine countersinking, necessitated by the use of flush-head rivets, produces a knife-edge in some of the GLARE plies. The resulting stress concentrations likely accelerate crack nucleation in such joints, diminishing fatigue performance. Furthermore, the superior fatigue performance of GLARE relative to monolithic aluminum permits reductions in skin thickness, increasing the risk of a through-thickness knife-edge.
This thesis examines the potential of dimple countersinking as an alternative method to machine countersinking in thin GLARE laminates. Using simple coupons, the dimpling process and crack initiation behaviour of dimpled GLARE is investigated. Additionally, wide lap joint fatigue specimens were tested to compare the relative performance of machine countersunk and dimpled GLARE lap joints. Initial results showed that dimpled GLARE joints have an inferior fatigue performance; however, these results are based on dimpling tools and processes optimized for aluminum. Delamination damage was observed during the dimpling process which may be responsible for the poor fatigue performance. Through optimization of the dimpling tool geometry, such damage may be avoidable, allowing the true potential of dimpling to be investigated.