Low cycle fatigue life improvement of AISI 304 by initial and intermittent wire brush hammering

Abstract

The effects of hammering by wire brush as a method of improving low cycle fatigue life of highly ductile austenitic stainless steel AISI 304 have been investigated through an experimental study combining imposed strain fatigue tests and assessment of surface characteristic changes under cyclic loading by SEM examinations and XRD analysis. It has been shown that the fatigue life of wire brush hammered surface was increased by 307% at an imposed strain rate of 0.2% and only 17% at an imposed strain rate of 0.5%, comparatively to the turned surface. This increase in fatigue life is explained in terms of fatigue damage that is related to crack networks characteristics and stability which are generated during fatigue on both turned and wire brush hammered surfaces. The improvement of brushed surface is attributed to the role of the surface topography, the near surface stabilized compressive residual stresses and superfi-cial cold work hardening on the fatigue crack network nucleation and growth. It is found that wire brush hammering produces a surface texture that favors, under cyclic loading, nucleation of randomly dispersed short cracks of the order of 40 lm in length stabilized by the compressive residual stress field that reached a value of r0 = ����749 MPa. In contrast, turned surface showed much longer unstable cracks of the order of 200 lm in length nucleated in the machining groves with high tendency to propagate under the effect of tensile residual stress field that reached value of r0 = 476 MPa. This improvement is limited to strain rates lower than 0.5%. At higher strain rates, a cyclic plastic deformation induced martensitic phase alters furthermore the fatigue behavior by producing high cyclic strengthening of the bulk mate-rial. This phenomenon lead to a reduction in strain imposed fatigue life. It has also been established that wire brush hammering can be used as an onsite surface treatment to improve the residual fatigue life of components subjected to cyclic loading. The efficiency of this treatment is demonstrated if it is performed at a fraction of service lifetime Ni/Nr lower than 0.5.