Research On Detection And Discrimination Of Rail Defect Based On Acoustic Emission Technique

In our country, the safe operating period of railway will replace of the large amount construction period. How to ensure a safe, stable and reliable operation will be an important problem in railway transportation. In rail defect detection, the conventional ultrasonic method is limited in detection speed and efficiency, and many researchers have struggled to find a new way to solve this problem. Be different from other detection methods, Acoustic Emission Technique(AET) is suited to detect dynamic crack, crack initiation and crack growth. Acoustic emission(AE) signals generated by cracks are able to reveal the crack nature, while the surface scanning of object for local defects is not required, and it only requires the receiving transducers. Therefore, it can be used to detect rail defects. However, little work has been done on studying the AE detection of rail defects, the possibilities of applying AE technique in rail detection have been only examined by theory and experiment, and many details need to be worked out. Based on the above situations, the detection and discrimination of rail defect based on AET are key points in this dissertation. The features of AE sources and analytical conditions, AE signal detection method at high speed, and discrimination rules of rail defects with different states are investigated. Major contributions and novelties of this dissertation can be summarized as following:The modeling method of AE by finite element is studied. In order to obtain the effective three-dimensional finite element model of AE, a moment tensor expression of crack source is given, and simulation method of AE is investigated. The AE propagation is simulated in two-dimensional space situation. Phase velocity curves, group velocity curves and propagation are examined to verify the effectiveness of described modeling methods. Based on group velocity curves of Rayleigh-Lamb equations and time-frequency image, modal analysis method of AE is given. These are of great significance for AE research in three-dimensional model of rail.The features of AE sources and analytical conditions of features in rail are studied. The research of AE sources in rail can help us to detect and identify crack sources effectively. The three-dimensional finite element model of rail with crack sources is created based on suitable modeling method of AE, and the effectiveness of this model is demonstrated by comparison of experiment results. A feature analytical method of AE sources in rail is proposed. Meanwhile, the features of AE sources in different source depths and different propagation distances are studied by simulation results. The validity of Rayleigh-Lamb wave equations in rail are investigated based on experimental method, and the application range is given. The influences of different measurement points on the signal features are analyzed and the choice method of measurement point is proposed. The results provide the features and analytical methods of AE sources with different types, different depths and different propagation distances. The analytical conditions of AE signals in rail are also proposed.The detection of AE signals at high speed are studied. A rail-wheel test rig with high speed is established based on Hertzian contact theory in order to study rail defect detection at high speed, the highest speed is 177km/h, and an experimental device is designed to make Hsu-Nielsen sources as a simulated acoustic emission source of rail defect. In order to obtain decomposition level and characteristic frequency range, the differences between noise signals and defect signals are analyzed in detail. Time-Shannon entropy is utilized to detect rail defect and the steps are also described. In order to suppress noise effects and ensure appropriate time resolution, the length of time window is investigated based on AE signal features. Further, the detection method of defect signals is proposed based on the features of time window at high speed. The experimental results demonstrate that the proposed method is effective for the detection of AE signals at high speed 124km/h.The discrimination rule of rail cracks with different states is studied. The states of rail steel from initial until final fracture are obtained by tensile testing, and AE signals are obtained by AE data acquisition system. According to the stress-time curve, AE signals with safe state and unsafe state are obtained. The features of AE signals in different crack states are examined based on AE activity and frequency spectrum. The relationships between AE features and different crack states are given in rail steel. Based on variation rate of AE hits and frequency features of AE signals, the detection criteria are established to detect the safe status of rail. After the combination of discrimination rule and detection method, an integrated method for rail defect detection and feature analyzing is completed, offering an effective way to detect and evaluate the safety status of rail.