Study On Crack Identification Of Slab Track Based On Transient Impact Response

The structure of slab track has a lot of damage in the complex service environment,which directly affects the safe operation of the train.As far as the orbital plate is concerned,there are many kinds of crack damage under the action of load.The existence of damage will change the dynamic characteristics of the structure.Failure to grasp the damage status and take treatment measures will cause accidents.Therefore,it is very important to identify the crack damage of slab track structure,and the most critical stage is to discriminate the degree of crack damage.(1)By using the self-made loading device to carry out the transient impact test on the track plate with crack damage,the arrangement of each loading point and the measuring points are studied,the feasibility of using transient impact method to compare and select the arrangement of measuring points is validated by analysis results of the test measurement,and provides the instruction for the finite element simulation.(2)The finite element model of slab track structure is established,and the different damage degree of crack is simulated by elastic modulus reduction.Based on the response signals of the measured points under transient impact,the time-frequency analysis method,the correlation degree analysis method and the transfer characteristic analysis method are used to study and obtain the characteristic indexes of the good mapping relationship with the crack damage degree.(3)In order to identify the damage effectively,reduce the information overlapping redundancy,the principal component analysis is applied to further optimize the characteristic indexes to determine the characterization parameters with good damage sensitivity and regularity,so as to the study of crack identification.(4)The damage characteristic parameter and the pattern recognition method of BP neural network are combined to identify the crack damage degree.The applicability and feasibility of the research method are verified by the response signal processing results of different fracture conditions under the test condition.