Fatigue crack growth in railway axles: Assessment concept and validation tests

On the residual life assessment of high-speed railway axles due to induction hardening

“…To consider the effects of stress ratios and the plasticity-induced crack closure (PICC), a crack opening factor f is introduced to modify the crack growth driving force parameters of ΔK eff and effective ΔK th,eff as [3,[7][8][9] (4) in which K max is the maximum SIF. Note here that only the PICC is quantitatively considered although the oxide-and roughness-induced crack closure might also produce a beneficial effect to enhance the crack closure load [3,[6][7][8][9][10]24]. Compared with standard NASGRO, it has been verified that the present LAPS can provide a slightly conservative residual life.…”

“…Ample studies has discussed the flaw morphology evolution and its characterization in both solid and hollow axles, a preferential semi-elliptical crack is well recognized with different aspect ratio a/c where a and c are the crack depth and length, respectively. It is therefore concluded that the aspect ratio is to be 0.4~0.7 usually for axles without surface residual stress [10]. However for those Shinkansen hollow axles or European deep rolled high-speed axles, compressive stress near the axle surface provides an increasing fatigue cracking resistance [1,4,13,14].…”

“…By contrast, standard NASGRO requires ample complex and expensive standard fracture mechanics experiments under different R, and then the artificially fitted parameters are optimized elaborately, leading to a considerable error of residual life of the damaged axles. Figure 2 shows the FCG predictions via LAPS and NASGRO compared with the experiment from companion specimens [10].…”

“…To acquire a real fatigue loading sequence for a reliable fracture mechanics assessment, a strain gauge far away from the press fits and fillets is elaborately conducted at the position of the journal surface with a linearly-distributed stress (see Figure 3). A typical five-stage block loads will be used for the lifetime of 34CrNiMo6 axles [10,25]. …”

“…However for full-scale axles, there are always pronounced deviations between predicted lifetimes and experiments even though a simulation model is validated. This well-known transferability due to evolved crack tip and stress field, load spectra, material behaviour, and geometry constraint is still open towards a better life design [7,[10][11][12]. Improving the fatigue crack resistance of the involved material can fundamentally help to lengthen the total service lifetime.…”