Strain Field Damage Dataset Construction Method Based on Finite Element Simulation

Deep learning shows great potential in the field of structural damage recognition, but training requires a large amount of high-quality data. In this paper, we take the subway sleeper girder as the research object, and use the finite element method to construct the strain field damage dataset. Based on the strain equivalence hypothesis to determine the damage simulation method, study the impact of damage units on the strain field of the sleeper beam, through the setting of different damage conditions for simulation and analysis, to verify the effectiveness of the method in the construction of the dataset, and to lay the foundation for the subsequent deep learning to realize the automated identification of structural damage of the subway sleeper beam.