Initial Study of Magnetic Induction Tomography for Crack Detection of Composite Pipe

Abstract

This study investigates the feasibility of Magnetic Induction Tomography (MIT) as a non-invasive method for detecting cracks in glass fiber-reinforced composite pipes, which are increasingly used in the oil and gas industry. An MIT system was modeled in COMSOL Multiphysics using eight copper coils arranged circumferentially around the pipe. Crack scenarios with varying opening angles, positions, and numbers were simulated. Sensor data were processed in MATLAB using a Linear Back Projection (LBP) algorithm to reconstruct tomographic images. The Mean Structural Similarity Index (MSSIM) values were generally below 0.1, with the highest value observed for a single crack with a 20° opening angle. The results indicate that MIT can detect and localize cracks in composite pipes, demonstrating the method’s potential for this application.