ABSTRACT
A new approach to combine stochastic simulation and pummelling analysis is proposed to better evaluate wheel/rail performance, which can be used for the optimization of wheel/rail characteristics and the prediction of track maintenance needs. This methodology was validated by comparing the results of the simulation with the data collected in the field. The pummelling results which mapped areas of rolling contact fatigue cracking were compared with areas of surface cracking seen in the photographs of the rails. The methodology was proved effective by good matches between the mapped areas of likely cracking and the cracked areas of the rails. Furthermore, the rolling contact fatigue crack orientations and positions seen on rail surface photographs were compared with the pummelling results to validate if the simulation had correctly predicted these crack characteristics.
