A series of experimental investigations concerning the residual stress fields at cold‐expanded fastener holes and of the behavior of fatigue cracks at such holes has been conducted. These studies have included measurement of the initial, cold‐work‐induced residual stress fields at both uncracked and cracked holes and the performance of both constant amplitude and spectrum fatigue crack growth tests.

A closed form solution technique first presented by Hsu and Fonnan for the radial expansion of a fastener hole is extended to include elastic-plastic unloading and therefore the creation of a reverse yielded zone. In addition, the dependence of interface pressure upon interference with an elastic insert is established. Numerous parametric studies are presented to demonstrate the effect of significant model variables and selected example calculations are compared with experimental results. NOMENCLATURE a = hole radius bl = radial distance to elastic-plastic boundary b, = radial distance to reverse yield boundary C = crack growth rate equation coefficient E =modulus of elasticity (plate) Ei = modulus of elasticity (insert) E, =secant modulus (plate) I, = applied expansion ratio 1, =interference ratio 1, = retained expansion ratio J2 = second invariant of the stress deviator tensor K,= crack growth rate equation coefficient n = strain hardening exponent (plate), crack growth rate equation exponent p = pressure r = radial coordinate in cylindrical coordinate system R = ratio of in-plane transverse plastic strain to through the thickness plastic strain, ratio cycle minimum to cycle maximum stress u = radial displacement B = Bauschinger parameter E = strain v = Poisson's ratio (plate) vi = Poisson's ratio (insert) 0 = tangential coordinate in cylindrical coordinate system u = stress ue = effective stress no = flow stress by = uniaxial yield strength

Notches or other stress concentrations are by far the most common sites for the initiation and growth of fatigue cracks in aircraft structures. The growth of these cracks is directly influenced by the material stress-strain response in the vicinity of the notch. Specifically, when the applied (remote) stress is sufficient to cause local plastic deformation at the notch, the response (local) stresses can no longer be found using elastic stress concentration factors, and they become dependent on the prior loading history. This is to say that the response stresses can no longer be treated as state variables. The occurrence of fatigue crack growth at notches which experience local yielding one or more times during their design lifetime is, in fact, quite common in many cyclically loaded structures. Some of the assumptions inherent in “traditional” Linear Elastic Fracture Mechanics (LEFM) based fatigue crack growth analysis may be inappropriate for such problems. In particular, the assumption that the stress distribution on a critical plane remains proportional to the elastic distribution throughout the loading history becomes incorrect when one or more of the applied loads causes plastic deformation and introduces or alters a residual stress field in this region. This paper first describes an elastic-plastic stress-strain response algorithm which may be used to estimate response stress distributions on a critical plane on a cycle-by-cycle basis. This is followed by a discussion of the manner by which stress intensity factors may be calculated based on these response stress distributions using Green's functions. Finally, the use of these stress intensity factors for the calculation of crack growth rate and, ultimately, crack growth life, is demonstrated.