Application of Machine Learning in Fracture Analysis of Edge Crack Semi-Infinite Elastic Plate

Abstract

This paper discusses the application of machine learning techniques, notably artificial neural networks (ANN), in the fracture analysis of semi-infinite elastic plates with edge cracks. The Stress Intensity Factor (SIF) model for a semi-infinite plate with a tip crack is employed in the study, and Finite Element Analysis (FEA) is performed via ABAQUS CAE to build a comprehensive dataset containing numerical simulations data. To improve accuracy and reliability, data preprocessing is implemented, and ANN as a valuable machine learning model is trained with various variables describing crack propagation, stress distribution, and plate structure as input parameters. The suggested method is compared to established fracture analysis methods, proving its accuracy in predicting crack behavior and stress distribution under a variety of loading circumstances. The model provides useful insights into the behavior of edge cracks in semi-infinite elastic plates, enhancing material engineering and structural mechanics. The study demonstrates the potential of combining FEA and machine learning to improve fracture analysis capabilities, and it discusses limitations and future research directions, encouraging the exploration of advanced machine learning techniques and broader fracture scenarios for future fracture mechanics innovation.