Evaluation of Fatigue Failure Risk Based on On-Track Tests

When a mechanical component of a railway vehicle experiences failure due to fatigue, it is necessary to determine whether it has been an isolated failure or it is a systematic failure that may affect the whole fleet. The present work presents the methodology to be followed for the application of on-track tests to determine the probability of fatigue failure. By simulating a normal commercial service in one or multiple conditions, the stress levels experienced by the component are recorded. These stress levels are combined and extrapolated in order to determine the typical global stress history that the component will experience during its operational life. From this stress history, the risk of fatigue failure can be estimated by means of the cumulative damage method also known as Miner’s rule. For this calculation the measured stress levels have to be compared with an S-N curve. The problem arises because the S-N curves presented in the standards correspond to stresses in the direction perpendicular or longitudinal to the weld. The peculiarities of on-track tests imply that the direction of principal stress may vary continuously and the maximum stress value may be present in specific points of the record with directions that do not represent the normal behaviour of the piece. Different forms of dealing with the projection of the rosettes have been analyzed, and the influence of these considerations on the damage calculation has been studied.