Fatigue failure of rivets and bolts in shear

Double-covered shear joints with hot rivets or with snug tight bolts subjected to cyclic load may fail in the ply or in the fastener. Tests with the latter failure mode are sparse and scattered. This paper combines these data and attempts to provide a unified theory on the fatigue driving force for this failure mode. Using the finite element method, we demonstrate that the nominal shear stress in the fastener is not a good indicator of its fatigue performance, because it ignores frictional force transfer and assumes an equal load share between fasteners. We provide an analytical model for the actual shear stress range. We derive S–N curves for this shear stress range, either using straightforward regression of test data or using Bayesian inference of the decisive failure type (ply or fastener). Finally, we derive the geometries and loads for which fastener failure is decisive over ply failure.