Probabilistic analysis of train–track vertical impact forces

Abstract

As part of a project developing a new limit states design method for prestressed concrete railway sleepers, extensive measurements were conducted of vertical impact forces exerted by trains on track. The vertical impact forces due to defects in the train wheels were found to occur randomly and followed an exponential distribution with time. This distribution is used to predict return periods for various levels of impact forces, needed for limit states design of sleepers. The effect of train speed was studied and showed that doubling the speed of a train would increase impact forces by about 140% for a wide range of return periods, but increasing the return period from 25 years to 200 years increased the impact force by only 10 to 15%. Applying a proposed cumulative damage model to the Monte Carlo method, it was found that the return period used for potential design of sleepers had little effect on the rate of sleeper failure in the period up to 20 years sleeper age, but that the return period had a large effect after that time.