Crack detection in bulk superconductor using Genetic Algorithm

Abstract

In this paper, we use the magnetic field and displacement distributions to detect crack in bulk superconductors. The (Re)BCO bulk superconductors are ceramic oxides which have low mechanical strength. During the field cooling magnetization, the large Lorentz force will be generated in bulk superconductor, which will cause stress concentration near the cracks. When the local stress exceeds the fracture toughness, cracks will propagate and even cause damage of the bulk. In order to ensure reliability of bulk superconductor, it is necessary to determine location and size of crack. For bulk superconductor with a crack, the crack can also affect the magnetic field distribution. Firstly, we combine the Genetic Algorithm (GA) and magnetic field to detect crack in bulk superconductor. The values of position and shape information of crack in bulk superconductor can be obtained with objective function. After that, we obtain the electromagnetic force based on the magnetic field distribution. The Extend Finite Element Method (XFEM) with electromagnetic force is used to calculate the displacement of the bulk superconductor with a crack. The displacement and GA are also used to detect crack in bulk superconductor. Finally, we present a comparison between the magnetic field detection and the displacement detection. For horizontal crack, the displacement detection is more effective than the magnetic field detection.