چكيده به لاتين
Fatigue and fracture of railway track components has become a critical problem recently which has caused high costs of maintenance and has consequently threatened the safety of mobility. On the other hand, in addition to casualties and financial loss, this phenomenon could lead to serious consequences.
Bolted rail joints are one of the sensitive components of railway tracks which are used in signaling and broken rail identification systems. Rail joints are the most vulnerable spots in the track structure on which many hours are spent for their maintenance. A common mode of failure in bolted rail joints is fatigue crack initiation and growth in the bolt holes at the end of the rail. Fatigue failure around rail-end bolt holes is particularly dangerous since it leads to derailment of trains and consequently causes inevitable accidents. However, crack growth has not been investigated in detail by researchers therefore in this thesis reliability analysis of the rail-end bolt hole crack growth is focused on.
Firstly, in this research a 3D finite element model of rail joint has been provided using commercially available software Abaqus/CAE. In this study, FE models are analyzed quasi-statically. After the verification of the model and checking the LEFM conditions, fracture mechanics analysis of cracks were carried out. At this stage, considering the stress intensity factor of cracks, the effects of variation of axle load, train speed and the rail support stiffness on the equivalent stress intensity factor were determined along with the ratio of KI to KII. In addition, step by step fatigue crack growth analysis has been conducted based on Paris law and Maximum Tangential Stress (MTS) criterion. As a result, an estimation of crack paths and cracks growth rate were obtained. In the next step, the quadratic model for equivalent stress intensity factor was described with the use of Response Surface Methodology (RSM) in terms of axle load, train speed, the rail support stiffness and crack length. In the end, the Monte Carlo Simulation (MCS) method was used for reliability analysis of rail-end bolt hole crack growth. The RSM model was used instead of FE model analysis in Monte Carlo simulations. In this analysis, limit-state function has been defined in term of loading cycles and the following parameters were considered as random variables: “the rail support stiffness”, “fracture toughness”, “cracks initial length”, “axle load”, “train speed”, “number of car body” and “number of trains passing per day”. Based on the results, 14 months after the cracks initiation, reliability index decreases rapidly from 5.7 to 0.0 due to the high rate of crack growth. Consequently, the risk of this deterioration should be considered. In addition, reliability index is more sensitive to train speed and axle load.
