Crack identification in magnetoelectroelastic materials using neural networks, self-organizing algorithms and boundary element method

• A hybrid approach using AI tools is proposed for crack identification in magnetoelectroelastic materials.
• A DBEM formulation is used to obtain the training set of a NN.
• The training set is separated into small training sets using only intrinsic properties of the data set, using the self-organizing algorithms.
• Extended displacements taken at a specific boundaries are sufficient to provide good damage identification.
• Noise sensitivity in the data set was reduced to a minimum using Gaussian mixtures algorithm as automated partitioning method.

In this paper, a hybrid approach that combines both supervised (neural networks) and unsupervised (self-organizing algorithms) techniques is developed for damage identification in magnetoelectroelastic (MEE) materials containing cracks. A hypersingular boundary element (BEM) formulation is used to obtain the solution to the direct problem (elastic displacements, electric and magnetic potentials) and create the corresponding training sets. Furthermore, the noise sensitivity of the resulting approach is analyzed. Results show that the proposed tool can be successfully applied to identify the location, orientation and length of different crack configurations.