PSI - Issue 23

Pavol Dlhý et al. / Procedia Structural Integrity 23 (2019) 185–190 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

189

5

Fig. 5 The non-slit ring specimen with highlighted marker lines for measurement.

Fig. 6 The slit ring specimen with highlighted marker lines for measurement.

Fig. 7 Tangential residual stress result from slith ring method determination with polynomial fitting through obtained data.

Fig. 8 Tangential residual stress result from slit ring method compared with data obtained from X-ray diffraction and neutron diffraction measurements.

depth from 10mm to 20 mm both diffraction methods agree together and present slightly higher residual stresses than slit ring method. This could be explained by higher scatter of the results obtained by slit ring method as it is visible in Fig.7. It is given by smaller deflection of the rings in this middle area, because residual stresses are reaching zero. In the tensile part of the graph (more than 30mm) again good agreement between slit ring method and X-ray diffraction is found. However, data obtained by neutron diffraction are shifted lower. It should be noted, that slit ring method and X-ray diffraction was measured on the surface of the slit ring, neutron diffraction technique used data from the centre of the ring. Also, with the depth of the ring there is strong change of the microstructure due to induction hardening technique used. Experimental determination of tangential residual stress in railway axle is presented. Procedure has been verified by FEM simulations and subsequently experimental measurement has been done. The obtained results are compared with the neutron diffraction method and X-ray diffraction to verify utility and precision of presented methodology. Results are generally in agreement with each other in compression part near surface however neutron diffraction shows smaller values of residual stress in tension part more than 30 mm deep under the surface. This area is not that important for lifetime prediction, therefore all three methods can be used for residual stress prediction. The next step will be development of the methodology for axial residual stress measurement. Then, it would be possible reconstruct the whole residual stress distribution in the railway axle and use this residual stress for lifetime prediction. 4. Conclusions