Citation

Easa SM. Transp. Res. A Policy Pract. 1994; 28(1): 1-15.

Copyright

(Copyright © 1994, Elsevier Publishing)

DOI

10.1016/0965-8564(94)90038-8

PMID

unavailable

Abstract

The existing method of sight distance design at railroad crossings is deterministic. This paper presents a probabilistic method that accounts for the variations of the component design variables and their correlations. Two cases are considered. Case 1 addresses the sight distance needs along the highway and the railroad for an approaching vehicle. Case 2 addresses the sight distance needs along the railroad for a stopped vehicle. The probabilistic method is based on the advanced first-order second moment of reliability analysis. Case 1 is modeled as a system with two parallel components. The probability of failure of each component and the system probability of failure are developed. Case 2 is modeled as a single-component system, and a design graph for sight distance needs along the railroad is presented. The method is validated using Monte Carlo simulation; its application is illustrated using numerical examples; and the sensitivity of the design variables is examined. The proposed method should result in safer operations at railroad grade crossings.