Simulation Analysis and Experimental Study on Pipeline Circumferential Crack Detection Based on Residual Magnetism Effect

Abstract

Abstract A method for crack detection on the surface of pipelines that uses residual magnetic detection technology based on the hysteresis effect after magnetizing the pipeline to magnetic saturation has been developed in this study. Support vector regression (SVR) was employed for quantitative assessment. This paper reviewed the magnetization process of the magnetizer, the generation of residual magnetism effect, and the detection of residual magnetic fields. An experimental and theoretical study examined how the crack size affects residual magnetic anomalies. The simulation results indicated that the characteristic values of the residual magnetic anomaly changed in a roughly linear fashion at a more significant crack depth. As the crack width increased, the variations in residual magnetic anomaly characteristic quantities were relatively mild. The input values were denoted by the extracted characteristic amounts. The width and depth parameters of the crack were used as output values. A trained SVR model was employed to estimate the crack size. The crack widths and depths were accurately determined at 62.5% and 85.42%, respectively by combining this detection method, SVR, and the experimental results. This study demonstrated the feasibility of using residual magnetic detection technology, combined with SVR, to detect circumferential cracks in pipelines, which is an essential supplement to internal magnetic flux leakage detection and assessing pipelines.