Abstract
In this paper, a series of experimental investigations was performed on full-scale railway axles to analyze the fatigue crack growth behavior of EA4T steel under load spectrum derived from real operating conditions. The experimental results were compared to life predictions carried out adopting two models: (i) the conventional Nasgro equation and (ii) the cyclic R-curve concept implemented in the Modified Nasgro equation for describing the crack growth behavior of an arbitrary crack length. The results show that the life predictions performed by means of the Modified Nasgro equation coincide well with the experimental results with an underestimation of the residual lifetime less than 32%, while the traditional Nasgro equation leads to significant overestimation (≈120%) of the residual lifetime for load spectra close to the in service scenario.
Highlights
Railway axles are critical components, and their design assessment procedure, as well as the maintenance program, require special attention
The experimental data are reported with blue dots, while the predictions performed with the Nasgro equation and the Modified Nasgro equation are indicated, respectively, with the green and red colors
The primary tool needed for reaching this objective is developing a robust residual lifetime prediction model which is applicable to realistic operational conditions of the wheelset
Summary
Railway axles are critical components, and their design assessment procedure, as well as the maintenance program, require special attention. The design route must consider the deterioration of the axles during service that can be promoted by (i) surface damage from ballast impacts [1,2,3], (ii) pitting from corrosion [4], and (iii) surface scars in press-fits [5,6,7]. All these types of damages can lead to fatigue crack growth under a sufficient level of local cyclic stresses. Establishing a quantitative procedure for determining residual axle lifetimes is of fundamental importance, and it is still the subject of scientific studies as demonstrated by several projects devoted to this issue, e.g., Widem [12], Euraxles [13], and Maraxil [14]
Published Version (Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have