PSI - Issue 57

Tobias Pertoll et al. / Procedia Structural Integrity 57 (2024) 250–261 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

251

2

1. Introduction Deep rolling is an effective and also historical commonly (Altenberger 2005) used mechanical post-treatment process applied on railway axles. During the process application, a deep rolling tool is pressed onto the rotating railway axle and at the same time moved with constant feed rate. In the contact area, the material plasticises in the region close to the surface, achieving three positive effects (Schulze 2006): the surface roughness is reduced, the material is strain hardened and, most importantly, compressive residual stresses are introduced (Delgado et al. 2016). These effects have a positive impact on the fatigue behaviour (Berstein and Fuchsbauer 1982; Regazzi et al. 2020; Saalfeld et al. 2019) and crack growth propagation behaviour (Delgado et al. 2016; Regazzi et al. 2014). The achievable effect of process strongly depends on the treated material and the applied deep rolling parameters. This investigation is conducted on a high-strength steel 34CrNiMo6 used for the manufacturing of railway axles. The material shows almost no change in hardness due to the deep rolling application (Pertoll et al. 2023a). Therefore, and according to the literature, it can be assumed that for this material the applied residual compressive stresses have the most significant effect on the resulting fatigue and crack propagation behaviour (Röttger et al. 2005; Altenberger 2005). The introduced residual stress distribution in-depth typically exhibits the following characteristics. Maximum compressive residual stresses at or slightly below the surface, followed by a reduction towards zero crossing and thus the transition to the tensile residual stress region. The zero crossing occurs in order to achieve a stress equilibrium within the component. The depth of zero crossing is a measure of the achieved effective depth of the post-treatment process. Deep rolling has advantages compared to other post-treatment processes, such as shot peening (Mahmoudi et al. 2016) or induction hardening (Gao et al. 2023), relating to the application in case of railway axles. This is mainly due to the achieved depth effect, the multiple achieved positive effects in one treatment and the economical and comparably easy integration in the manufacturing process. Years of operation in the field can cause defects on the surface, e.g. due to ballast impact or corrosion (Gao et al. 2022; Hu et al. 2021). These defects can subsequently lead to (fatigue) cracks in the component. Several studies have already been published on crack propagation behaviour and the definition of maintenance intervals of railway axles. A sound overview is presented in (Zerbst et al. 2013). In general, the starting crack depth which can be reliably detectable by means of non-destructive crack testing methods exhibits a size of 1-2 mm (Richard and Sander 2012; Carboni and Beretta 2007). Some experimental investigations have been carried out on full scale axles (Rieger et al. 2020; Pourheidar et al. 2021) and on small scale samples (Simunek et al. 2020) with introduced initial crack notches. In addition to the experimental investigations, there are various simulation models focussing on the crack growth behaviour. Examples of analytical calculation tools for the calculation on railway axles are the software ERWIN (Lütkepohl et al. 2009) and INARA (MCL 2023), which is utilized in this study. An example for a software on numerical basis is Franc3D (Fracture Analysis Consultants 2023). The number of investigations, experimental or numerical, on the crack behaviour of deep rolled railway are generally limited. For example, (Regazzi et al. 2014; Hassani-Gangaraj et al. 2015). (Hassani-Gangaraj et al. 2015) showed that the introduced residual stresses and thus also the crack propagation behaviour strongly depend on the process parameters and thus there is also potential to optimise the process application in this respect. Hence, this study focusses on the influence of the residual stress states introduced by deep rolling on the resulting crack propagation characteristics. Therefore, a validated numerical deep rolling simulation model is used to numerically evaluate the residual stress distribution in the near-surface region of the railway axle after deep rolling. These results are subsequently processed applying the elaborated crack growth software tool INARA to calculate the crack propagation behaviour of railway axles. This approach enables the investigation of the influence of the most influential deep rolling parameter, the deep rolling force, on the crack propagation of defective railway axles loaded with constant as well as in-service variable load amplitudes. The content and scientific contribution of this publication can be summarised as follows: • Investigation of the influence of the most significant deep rolling parameter, deep rolling force, on crack propagation behaviour • Determination of the critical threshold crack depth, considering various deep rolling forces and initial crack depths and geometries.