Research On Key Technology Of Non-destructive Detection Of Railway Train Axle Damage

With the continuous development of China's high-speed railway train,increasing the number of heavy haul train and growth of train running cycle,the train axle will be subjected to complex external force.Train axle works in high cycle,high strength conditions.The material of axle prone to generate fatigue cracks and cracks location hidden,under the action of torsion,stretching and compression,when the crack size to a certain length,there may be a shaft fracture accident if the train axle is detected untimely,it will bring huge losses and disasters to national property and people's security.It is necessary for train axle to carry on non-destructive testing,and find defect timely in preventive measures.This paper aims to study the related technology of phased array ultrasonic axle detection.Based on a specific train axle detection platform,using the characteristics of phased array ultrasonic sound beam deflection,focusing and deflection focusing,Designing a axle detection strategy of "comprehensive scan&key flaw detection" under the conditions of single probe.In this paper,the phased array ultrasonic emission and reception system is designed.The emphasis is on the design of the 8-channel phased array ultrasonic emission timing control module and the improvement of the existing high voltage pulse transmission circuit.The program of phased array ultrasonic beam deflection and beam focusing is written in Verilog language,and the timing control module is simulated by Model Sim software.The rationality of the designed high voltage pulse transmission circuit is verified by Proteus software.In this paper,the ME model is used as the axle phased array ultrasonic flaw echo signal model.Based on this model,the wavelet denoising of the ultrasonic echo signal is studied.Discussing the issues of the optimal wavelet basis,the number of decomposed layers and the corresponding noise reduction threshold.The optimal wavelet basis,the number of decomposed layers and the corresponding noise reduction threshold are determined by calculation and MATLAB simulation.Based on the noise reduction of ultrasonic echo,the defect echo signal is analyzed from the time domain,frequency domain and time-frequency domain respectively.The echo signal is decomposed by wavelet packet,and different defects eigenvalues are extracted,and constructing eigenvectors achieves the purpose of identifying defects.Then,the location and quantification of the defect are studied based on the specific axle defect detection platform.The distance between the end of the axle defect is determined by the encoder on the detection platform.The time point of the defect signal is found by A scanning time domain.It is known that the ultrasonic wave propagates in the steel speed and time can calculate the depth of the buried buried.Finally,we introduce the method of calculating the size of defect by using-6d B method and curve tracking method.