Simulation Research on Damage Identification of Plate Weld Based on Correlation Vector Amplitude Vector

Abstract

A method for detecting the quality of the plate weld based on the cross-correlation function magnitude vector is proposed. In the large-scale plate welding process, the weld is the most vulnerable part of the damage. This paper proposes a method to accurately determine the location of the weld defect. In this paper, the finite element software ABAQUS is used for numerical simulation to establish a plate weld model, and the weld area is divided into 19 units. In order to avoid noise interference, a 10 second 1000N stationary random excitation is used, and 10 excitation reference points are arranged around the random excitation point, and the best reference point is determined by the simulation result. The damage of each weld unit is characterized as the decrease of Young’s modulus, the acceleration response is collected at the joint of each weld unit, and the single defect, double defects and multi-defects conditions are simulated respectively. The numerical results show that the cross-correlation amplitude vector method can always accurately locate the defects under various damage conditions.