Railway track comprises of continuous welded rail mounted with rail clips on sleepers integrated to a ballast track form system. Modeling the rail structure with thermite weld subjected to the complex dynamic loadings is a challenging problem. Fatigue failures at the head‐to‐web, web‐to‐foot, and foot regions of weld collar are investigated. In this paper, a combined method of multibody system dynamic analysis and dynamic finite element analysis was used. A train roll‐in experiment was conducted at a train depot test track to validate modeling results predicted at the strain gauge location. Three critical plane‐based multiaxial fatigue criterions incorporated with a smallest enclosing circle algorithm were implemented in Python code to study the fatigue behavior at weld collar. Parametric studies were also conducted to investigate the effects of track component materials, track curvature, and train velocity. This approach provides a method for predicting the failures of thermite welded joints in railway tracks.

No abstract

No abstract