Resumen
In this paper, the fatigue design of rails in the transition from bridge to abutment is examined. The current fatigue design of rail systems is carried out with simplifying assumptions concerning the stresses in the rail. Thus, the determination of a fatigue endurance limit is done with stresses based on the Smith-Diagram. By applying this means of design, the assumption is made that there are mainly normal stresses in a rail. Those may originate from changes in temperature, accelerating or slowing down of the trains as well as bending stresses caused by the passing of trains. Horizontal components of the load (from the train's sine run, lateral shift of bridges or the cant of curves) are not considered in this design format. At the same time, specimens in fatigue tests with pulsing vertical loads regularly show unexpected and inexplicable cracks at the edge of the rail foot and a huge scatter of the results. In cases of rail breaks the location of the incipient crack is not necessarily in the cross section's area with the largest tensile stresses, but can be shifted to areas of smaller tensile stress (e.g. the edge of the rail foot). Obviously additional criteria such as the condition of the surface of the rail foot in combination with stresses from warping of the section (secondary bending) have more influence on the fatigue performance than estimated before. This paper presents an approach to evaluate the importance and priorities of the named influencing factors and criteria on the fatigue design of rails by means of fracture mechanics. The focus hereby lies on the rail foot, since in this region of the cross section there are no disturbing influences as contact problems or wear.