Rail failure is one of the serious problems in the railway industry. Rail failures can significantly impact rail safety as well as foster numerous maintenance challenges. A better understanding of the parameters that lead to rail failures and the ability to predict their uncertainties would result in the facilitation of a more focused maintenance procedure. Rail under-head failure (RUF) occurrence is common in heavy haul rail operations and hence chosen for the modelling and discussions presented in this paper. Stochastic risk analysis of RUF was conducted to determine the probability of its occurrence. The parameters included in the analysis are: vertical elastic foundation of the rail, daily and/or seasonal temperature variances, headwear, a contact patch offset (CPO) and the lateral to vertical force ratio (L/V) acting on the rail head. The methodology involves both qualitative and quantitative risk analyses of the above failure parameters. The qualitative analysis was performed by using fault tree analysis and Boolean algebra to determine minimum cut sets. Further, the failure parameter probabilities for the developed quantitative model are assigned after consulting expert opinion and previous research. A simple quantitative model using triangular distributions was created to serve as a template to perform more complex quantitative modelling with higher level probabilistic distributions. The results were analysed and compared with existing research. The results indicated 14-23.5% probability for RUF occurrence in 90% of the cases. Both the simple and complex models had very similar results and displayed skewness’s that match previous infrastructural studies on risk analysis. The results conclude that appropriate actions should be taken in maintenance planning to prevent the parameters that can lead to RUF.

Palaneeswaran Ekambaram

Affiliation

Conference Tracks

Refine Results

Stochastic Modeling of Rail Under-head Failure Risks