Predicting Dynamic Force Level in Railway Crossings

Railway wheels negotiating a crossing are known to generate a dynamic impact load while transferring from the wing rail to the nose and vice versa. Dynamic impact loads have been studied in the 70s in the UK and the so called P2 analytical formula derived against site measurement for the case of rail joint. The same formula has then been adapted to the case of crossing and used in the UK to analyze crossing performance against foot fatigue resistance. However, the original formula has not been thoroughly verified for the case of crossings and it is found to differ from the more detailed MBS based prediction. The work presented here is reviewing the original work by Jenkins from which the P2 formula originates and establishes how it compares with respect to a number of more advanced options to simulated wheels unsprung mass in the presence of a rail dip angle (rail joint or crossing) to advise on future revision of the formula better adapted to the case of crossing. Overall it is shown that most numerical models are providing consistent results but that the Jenkins formula is over estimating the loads, especially above 200 km/h and as the quantity speed times dip angle increase. Track models based on beam with discrete support provide more accurate prediction of detailed interaction with passing loads, that other simplified co-running track models cannot.