Estimation of multiple cracks interaction and its effect on stress intensity factors under mixed load by artificial neural networks.

One of the goals of fracture mechanics in the analysis of problems involving cracks is to investigate crack growth and propagation. Each crack grows and propagates in the environment based on its geometry and loading conditions, which can occur under pure mode or a combination of modes. The growth and propagation pattern of the crack in these conditions aligns with the fracture mode. In the presence of multiple cracks in the environment, the growth and propagation paths of these cracks differ from the isolated crack case, where they can either grow independently under certain conditions or merge together. Nowadays, extensive research has been conducted on cracks and their mathematical relationships with respect to the geometry and loading conditions of various components. Since in many cases, a large number of cracks may occur in proximity to each other, the intensification and protective effects of each crack on others and their impact on the overall behaviour of components are of great importance. This study focuses on investigating the behaviour of interaction and the stress intensity factor for an infinite plate with parallel adjacent cracks. Remote stress boundary conditions in the form of tensile and compressive stresses have been applied to this plate, and using the J-integral and the finite element method with the Abaqus finite element analysis software, this subject has been examined. The obtained results indicate that depending on the type of vertical and horizontal distances and the angle of their placement between the adjacent crack and the main crack in the first, second, and mixed modes, the adjacent crack can have intensifying, protective, or neutral effects. Furthermore, a good estimation of the secondary crack's influence on the main crack has been provided using a neural network method (MLP), which creates an appropriate weight function for the upcoming sample. [ABSTRACT FROM AUTHOR]