PSI - Issue 4

D. Simunek/ Structural Integrity Procedia 00 (2017) 000 – 000

4

David Simunek et al. / Procedia Structural Integrity 4 (2017) 27–34

30

Fig. 3 Functionality of 1:3 scale specimen rotary bending test rig

2. Experimental investigations

Constant amplitude and overload tests are performed to compare crack propagation for different specimens. First, a crack initiation stage is performed and after a subsequent load reduction stage, a constant amplitude test load is applied. In Luke et al. (2011) constant amplitude and overload tests with different overload factors R OL are reported and retardation effects are analysed. The results highlight the possibility of transferability for crack growth parameters.

2.1. Constant amplitude (CA) tests

The stress intensity factor range ∆K varies along the crack front. Hence, three points on the semi-elliptical crack front are of particular interest for the crack growth assessment; the crack depth a 0 and the two surface points of the crack tip. Many papers deal with the behavior of stress intensity and geometry factors in railway axles and round bars under rotary bending (see Carpinteri et al. (1996), Carpinteri et al. (1998), Madia et al. (2008) and Madia et al. (2011)). The geometry factor of the 1:3 scale specimens is based on work in EBFW 2 (Lütkepohl et al. (2009)) and is implemented in the software INARA (Integrity Assessment for Railway Axles), which is an analytical assessment tool. With the help of this software package the stress intensity factor range is determined in dependence of the crack length and the a/c-ratio. Based on these calculations and the measured crack length during testing, the crack propagation diagram is generated. The evaluation of the constant amplitude tests on 1:3 scale specimens are presented in Figure 4. The results show a comparatively low, but still clearly observable, scatter of the test data. Compared to the specimens EA4T 2, 3 and 4, EA4T 1 shows a slightly slower crack propagation rate resulting in a higher residual lifetime, see Figure 4. One explanation for the observed scatter may be explained by different residual stress conditions, which affect the threshold value, further fatigue crack growth and the final lifetime significantly. Further investigations by residual stress measurements are scheduled to investigate this behavior in detail.