PROBABILISTIC STRUCTURAL ANALYSIS OF RAILROAD SUBGRADE USING FINITE ELEMENT ANALYSIS

摘要

The reliability of finite element (FE) based deterministic structural analysis of railroad trackbed has improved significantly due to ever increasing computational powers. However, the application of probabilistic methods to incorporate the material and geometric variabilities in these FE analyses is still profoundly underworked. In this study, the influence of variability in granular layers' modulus and thicknesses values on the railroad subgrade performance have been investigated by applying probabilistic method and using a 3-dimensional FE based numerical trackbed model previously developed and validated by the authors' research group. The influence of these factors is accounted for by changing their coefficients of variance (COV) while keeping their means constant. Preliminary results revealed that the variation in subgrade modulus is the most influential factor for subgrade performance, both in terms of progressive shear failure and excessive plastic deformations, followed by ballast modulus. Variations in depths, for the range studied, remained passive to the subgrade performance. The findings of this work is of particular significance in evaluating the subgrade performance while including the material and geometric variations, which may be caused by construction imperfections, weather changes, and/or rail operations.