Rail Corrugation Growth on Curves

Doktorsavhandling, 2012

The development of periodic irregularities with distinct wavelengths (corrugation) on the low rail of small radius curves is studied through mathematical modelling, numerical simulations, field measurements and laboratory investigations. One year of monitoring of roughness on the low rail of a 120 m radius curve on the metro of Stockholm Public Transport (SL) showed severe growth of rail corrugation with wavelengths of about 5 cm and 8 cm. About 300 days after rail grinding, the corrugation was observed to reach a constant amplitude. Based on a section removed from the corrugated rail, a laboratory investigation showed plastic deformation in the lateral direction towards the field side. No significant difference in microstructure was found when corrugation troughs and peaks were compared. A time-domain model for the prediction of long-term roughness growth on curves has been developed and validated versus field measurements. The simulation model is able to simultaneously capture the low-frequency vehicle dynamics due to curving and the high-frequency dynamics due to excitation by for example short-pitch corrugation. Non-Hertzian and non-steady effects in the wheel‒rail contact are considered. Simulations show that the short-pitch corrugation on the small radius curve at SL is generated by wear induced by the leading wheelset of passing bogies. The corrugation wavelengths 5 cm and 8 cm are determined by the excitation of the first antisymmetric and first symmetric bending eigenmodes of the wheelset, respectively. The importance of accounting for the phase difference between the calculated wear and the present rail irregularity in predictions of corrugation growth is demonstrated. Due to a phase difference approaching a low constant value, the growth of corrugation is predicted to eventually develop into a stationary state where it is translated along the rail with a constant amplitude. For track geometry and traffic conditions corresponding to the selected curve at SL, simulations indicate the wheel–rail friction coefficient to have a significant influence on corrugation growth. For friction coefficient 0.6 (measured at dry contact conditions), corrugation growth is predicted at several wavelengths whereas for friction coefficient 0.3 (due to application of a friction modifier) it is shown that an initial rail irregularity is gradually worn off by passing traffic. Based on a new set of field measurements in the same curve (another year of monitoring), it was shown that the application of a friction modifier directly after grinding is an effective mitigation measure to prevent the development of rail corrugation.